Final_V3_CLV_with_regression-14092023-FINAL.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "e0Ura4zS0xa4"
   },
   "source": [
    "<h1>CM3015 Machine Learning And Neural Networks - End of Term Coursework</h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "WimtGoUf0wqa"
   },
   "source": [
    "This machine learning excersie follows the follows the 'The universal workflow of machine learning' process from François Chollet's book, 'Deep Learning with Python' (second edition)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "1WrBxh6H01_Q"
   },
   "source": [
    "<h2> 1. Defining the problem and assembling a dataset </h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Pp7viPN-08FF"
   },
   "source": [
    "<h3>1.1 Frame the problem</h3>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2EZXfMob0_ZX"
   },
   "source": [
    "<h3>Customer research with Machine Learning</h3>\n",
    "\n",
    "I would like to learn more about the customers' behaviour and try to predict CLV with machine learning regression:\n",
    "\n",
    "<strong>Customer Lifetime Value (CLV):</strong> Estimate the future value of customers over their lifetime based on their historical spending patterns and interactions with the company\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "BpHoFKIw1OqO",
    "outputId": "663fb2ee-8a12-4a77-f499-9d5104a1fc58"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.13.0\n"
     ]
    }
   ],
   "source": [
    "# Use seaborn for pairplot.\n",
    "!pip install -q seaborn\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import tensorflow as tf\n",
    "\n",
    "from tensorflow import keras\n",
    "from keras import layers\n",
    "\n",
    "print(tf.__version__)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "zvk53gct1AGJ"
   },
   "source": [
    "<h3>1.2 Collect the dataset</h3>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "XSoPgtlP1b3e"
   },
   "source": [
    "The dataset I use is from kaggle.com.\n",
    "\n",
    "It contains information about visitors, visits and purchases from a shop that has a store, a website and a catalog.\n",
    "The data includes basic information about the customer, such as birth year, education level, marital status or household income, etc.\n",
    "It also contains information about their purchasing behaviour, such as how much they had spent on different categories (Wine, Meat, Sweets, etc.) and how many times they purchased through different channels (store, web, catalog).\n",
    "\n",
    "Source: https://www.kaggle.com/datasets/imakash3011/customer-personality-analysis\n",
    "\n",
    "I also used this tutorial to get a practical example of regression with TensorFlow:\n",
    "https://www.youtube.com/watch?v=-vHQub0NXI4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "uOlzqD401bKN",
    "outputId": "8d5359ee-8230-4abc-b001-d1edccacae82"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     ID  Year_Birth   Education Marital_Status    Income  Kidhome  Teenhome  \\\n",
      "0  9432        1977  Graduation       Together  666666.0        1         0   \n",
      "1  1503        1976         PhD       Together  162397.0        1         1   \n",
      "2  1501        1982         PhD        Married  160803.0        0         0   \n",
      "3  5336        1971      Master       Together  157733.0        1         0   \n",
      "4  8475        1973         PhD        Married  157243.0        0         1   \n",
      "\n",
      "  Dt_Customer  Recency  MntWines  ...  NumWebVisitsMonth  AcceptedCmp3  \\\n",
      "0    28.08.20      830         9  ...                  6             0   \n",
      "1    21.04.20      474        85  ...                  1             0   \n",
      "2    23.02.20      169        55  ...                  0             0   \n",
      "3    08.11.22      196        39  ...                  1             0   \n",
      "4    31.08.20      860        20  ...                  0             0   \n",
      "\n",
      "   AcceptedCmp4  AcceptedCmp5  AcceptedCmp1  AcceptedCmp2  Complain  \\\n",
      "0             0             0             0             0         0   \n",
      "1             0             0             0             0         0   \n",
      "2             0             0             0             0         0   \n",
      "3             0             0             0             0         0   \n",
      "4             0             0             0             0         0   \n",
      "\n",
      "   Z_CostContact  Z_Revenue  Response  \n",
      "0              3         11         0  \n",
      "1              3         11         0  \n",
      "2              3         11         0  \n",
      "3              3         11         0  \n",
      "4              3         11         0  \n",
      "\n",
      "[5 rows x 29 columns]\n"
     ]
    }
   ],
   "source": [
    "#read the data from file\n",
    "df = pd.read_csv('sample_data/ecommerce2.csv')\n",
    "\n",
    "print(df.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "6-H1XbpZ1qwo"
   },
   "source": [
    "<h3>1.3 Understand the data</h3>\n",
    "\n",
    "According to the book, before we \"start training models,\n",
    "we should explore and visualize our data to gain insights about what makes it\n",
    "predictive, which will inform feature engineering and screen for potential issues.\"\n",
    "\n",
    "1. Plot histogram of feature values to get a feel for the range and frequency of different values.\n",
    "2. If there are missing values, we have to deal with it\n",
    "3. Print the number of instances of each class in our data\n",
    "4. Check for target leaking"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Qa4izvjZ1zXd"
   },
   "source": [
    "<h4>1.3.1 Plot histograms</h4>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "5olhymXs187Z"
   },
   "source": [
    "I plotted some histograms just to have a generic overview of the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 432
    },
    "id": "wa8XCSv_17mM",
    "outputId": "d5950090-0cda-462a-f27a-9e9a0a5831cf"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABdEAAAHqCAYAAADrpwd3AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAACFTUlEQVR4nOzdf3zN9f//8ft+2A9jm5HN3piFYn5HGEqxGq3efvVD78X86Je28qOId+V3JhVSIqVR8VYq/aBkjXi/M7+GEkKFKTYV21Bmtuf3j75eH8d2ss3Zzn7crpfLuXBer+fr9Xo8n+ecPc55vF7neVyMMUYAAAAAAAAAACAfV2cHAAAAAAAAAABAWUURHQAAAAAAAAAAOyiiAwAAAAAAAABgB0V0AAAAAAAAAADsoIgOAAAAAAAAAIAdFNEBAAAAAAAAALCDIjoAAAAAAAAAAHZQRAcAAAAAAAAAwA6K6AAAAAAAAAAA2EERHShBDRo00KBBg5wdRoX3/PPP6+qrr5abm5tat27t7HAAAKWAHFs6SivHLlq0SC4uLjp06FCJHeNKTZw4US4uLs4OAwDKBPJw6Shvn3VLOlceOnRILi4ueuGFF4oVS2Gft+XhfQlKH0V0oJAu/BHdtm1bgetvuukmNW/e/IqP89lnn2nixIlXvJ/KYs2aNRozZow6d+6shIQETZs2rVDb3X333XJxcdGTTz5ZwhECAC6HHFs2FSXHDho0SC4uLgXevLy8SjHqK/PHH39o4sSJ+uqrr5wdCgCUGvJw2VSYPPzVV1/Zzb+X3i7n6NGjmjhxonbu3FkCvbm8nTt36r777lO9evXk6empgIAARUREKCEhQbm5uU6JCbiYu7MDACqyffv2ydW1aOeqPvvsM82dO5c3F4W0du1aubq6auHChfLw8CjUNllZWfr000/VoEED/ec//9H06dO5sgwAyhlybMkrao719PTUG2+8kW+5m5tbSYRXIv744w9NmjRJ0l9Fo4s9/fTTGjt2rBOiAoCyhzxc8gqTh5s2baq3337bZtm4ceNUrVo1PfXUU0U63tGjRzVp0iQ1aNCg1K96f+ONN/Twww8rMDBQAwYMUOPGjXXq1CklJSVp6NChOnbsmP79738XaZ/kbTgaRXSgBHl6ejo7hCI7c+aMfHx8nB1GoR0/flze3t6FLqBL0gcffKDc3Fy9+eab6tatmzZs2KCuXbuWYJQAAEcjx5a8ouZYd3d33XfffSUclfO4u7vL3Z2PTwAgkYdLQ2HycGBgYL7cO336dNWqVavc5ORNmzbp4YcfVnh4uD777DNVr17dWjdixAht27ZN3333XZH3W9J52xijs2fPytvbu8SOgbKF6VyAEnTpfFs5OTmaNGmSGjduLC8vL9WsWVNdunRRYmKipL++Cj137lxJKvBrV2fOnNHjjz9ufb3p2muv1QsvvCBjjM1x//zzTz322GOqVauWqlevrn/+85/65Zdf5OLiYnPW/8IcYXv27NG//vUv1ahRQ126dJEkffvttxo0aJCuvvpqeXl5KSgoSEOGDNHvv/9uc6wL+9i/f7/uu+8++fn56aqrrtIzzzwjY4yOHDmiXr16ydfXV0FBQXrxxRcLNXbnz5/XlClT1LBhQ3l6eqpBgwb697//rezsbKuNi4uLEhISdObMGWusFi1adNl9L1myRLfccotuvvlmNW3aVEuWLCmw3bfffquuXbvK29tbdevW1dSpU5WQkFDg3Giff/65brjhBvn4+Kh69eqKiorS7t27C9VXAEDRkWPLZo4tjN27d6tbt242+TUvLy9fu0vH9IKC5jPNyMjQyJEj1aBBA3l6eqpu3boaOHCgfvvtN0nSuXPnNH78eLVt21Z+fn7y8fHRDTfcoHXr1ln7OHTokK666ipJ0qRJk6x+X4ihoLlVCzOWF2K+/fbb9b///U/t27eXl5eXrr76ar311ltFHT4AKBPIw+UrD//000+66667FBAQoKpVq6pjx45atWqVtf6rr77S9ddfL0kaPHhwvmP+97//1V133aX69evL09NT9erV08iRI/Xnn38WO6YLLuTcJUuW2BTQL2jXrl2B85gvWLDAGsPrr79eW7dutVlf2PnZC/u+5EIu/+KLL9SuXTt5e3vrtddek/TX+5ARI0ZYz99GjRrpueees9nPxfO5Xy52lE1cSgEUUWZmpvWB7GI5OTmX3XbixImKj4/X/fffr/bt2ysrK0vbtm3T9u3bdcstt+ihhx7S0aNHlZiYmO8rWcYY/fOf/9S6des0dOhQtW7dWl988YVGjx6tX375RbNmzbLaDho0SO+9954GDBigjh07av369YqKirIb11133aXGjRtr2rRp1puUxMRE/fTTTxo8eLCCgoK0e/duLViwQLt379amTZvyJaN77rlHTZs21fTp07Vq1SpNnTpVAQEBeu2119StWzc999xzWrJkiZ544gldf/31uvHGG/92rO6//34tXrxYd955px5//HFt3rxZ8fHx2rt3r1asWCFJevvtt7VgwQJt2bLF+vp4p06d/na/R48e1bp167R48WJJ0r333qtZs2bplVdesTnD/8svv+jmm2+Wi4uLxo0bJx8fH73xxhsFXnHx9ttvKyYmRpGRkXruuef0xx9/aN68eerSpYt27NihBg0a/G1MAIC/kGPLd46VVODj5+HhIV9fX0lSWlqabr75Zp0/f15jx46Vj4+PFixYcEVXcZ0+fVo33HCD9u7dqyFDhui6667Tb7/9pk8++UQ///yzatWqpaysLL3xxhu699579cADD+jUqVNauHChIiMjtWXLFrVu3VpXXXWV5s2bp2HDhqlPnz7q27evJKlly5Z2j12Ysbzghx9+0J133qmhQ4cqJiZGb775pgYNGqS2bduqWbNmxe4/ADgKebj85+GCpKenq1OnTvrjjz/02GOPqWbNmlq8eLH++c9/6v3331efPn3UtGlTTZ48WePHj9eDDz6oG264weaYy5cv1x9//KFhw4apZs2a2rJli15++WX9/PPPWr58ebHikv6aRi0pKUk33nij6tevX+jtli5dqlOnTumhhx6Si4uLZsyYob59++qnn35SlSpVCr2for4v2bdvn+6991499NBDeuCBB3Tttdfqjz/+UNeuXfXLL7/ooYceUv369bVx40aNGzdOx44d0+zZs0skdjiBAVAoCQkJRtLf3po1a2azTUhIiImJibHut2rVykRFRf3tcWJjY01BL82PPvrISDJTp061WX7nnXcaFxcX88MPPxhjjElJSTGSzIgRI2zaDRo0yEgyEyZMsJZNmDDBSDL33ntvvuP98ccf+Zb95z//MZLMhg0b8u3jwQcftJadP3/e1K1b17i4uJjp06dby0+ePGm8vb1txqQgO3fuNJLM/fffb7P8iSeeMJLM2rVrrWUxMTHGx8fnb/d3sRdeeMF4e3ubrKwsY4wx+/fvN5LMihUrbNo9+uijxsXFxezYscNa9vvvv5uAgAAjyRw8eNAYY8ypU6eMv7+/eeCBB2y2T0tLM35+fvmWAwDyI8eW/xwbExNj97GLjIy02o0YMcJIMps3b7aWHT9+3Pj5+dnkV2NMvjG94NLHfvz48UaS+fDDD/O1zcvLM8b8NW7Z2dk2606ePGkCAwPNkCFDrGW//vqr3eNeeDwuKMpYhoSE5Ht8jx8/bjw9Pc3jjz+e71gAUJrIw+U/D1+sWbNmpmvXrtb9C7n3v//9r7Xs1KlTJjQ01DRo0MDk5uYaY4zZunWrkWQSEhLy7bOgMYuPjzcuLi7m8OHD1rJLc+XlfPPNN0aSGT58eKHaHzx40EgyNWvWNCdOnLCWf/zxx0aS+fTTT/82lkuft0V5X3Ihl69evdpmn1OmTDE+Pj5m//79NsvHjh1r3NzcTGpqapFjR9nEdC5AEc2dO1eJiYn5bn93ldIF/v7+2r17tw4cOFDk43722Wdyc3PTY489ZrP88ccflzFGn3/+uSRp9erVkqRHHnnEpt2jjz5qd98PP/xwvmUXn3k9e/asfvvtN3Xs2FGStH379nzt77//fuv/bm5uateunYwxGjp0qLXc399f1157rX766Se7sUh/9VWSRo0aZbP88ccflySbr50V1ZIlSxQVFWV9Taxx48Zq27ZtvildVq9erfDwcJsfVAkICFB0dLRNu8TERGVkZOjee+/Vb7/9Zt3c3NzUoUMHm6+JAwD+Hjm2fOdYLy+vAh+/6dOn2xy/Y8eOat++vbXsqquuypdfi+KDDz5Qq1at1KdPn3zrLlxN6ObmZn3jLC8vTydOnND58+fVrl27Ase8MIo6lmFhYdaVfdJf/S7MYwYApYU8XL7z8N8ds3379tZ0NpJUrVo1Pfjggzp06JD27Nlz2X1cPGZnzpzRb7/9pk6dOskYox07dhQ7tqysLEkqcBqXv3PPPfeoRo0a1v0L+bWoObWo70tCQ0MVGRlps2z58uW64YYbVKNGDZuaQEREhHJzc7Vhw4YSiR2lj+lcgCJq37692rVrl2/5hT+Yf2fy5Mnq1auXrrnmGjVv3lw9evTQgAEDCvWm5PDhwwoODs6XXJo2bWqtv/Cvq6urQkNDbdo1atTI7r4vbStJJ06c0KRJk7Rs2TIdP37cZl1mZma+9pd+9crPz09eXl6qVatWvuWXzjV3qQt9uDTmoKAg+fv7W30tqr1792rHjh0aOHCgfvjhB2v5TTfdpLlz5yorK8v6uvnhw4cVHh6ebx+XxnThTWK3bt0KPOaF/QEALo8cW35zrPRXYSEiIuKyx+/QoUO+5ddee22xj/vjjz+qX79+l223ePFivfjii/r+++9tpiYo6DEqjKKOZUFfU69Ro4ZOnjxZrOMDgKORh8t3Hv67YxaUey8e3+bNm//tPlJTUzV+/Hh98skn+fJWQWNWWBc+L586dapI2136mFwoShc1pxb1fUlBz6cDBw7o22+/tX5X5VKXPsccFTtKH0V0oBTdeOON+vHHH/Xxxx9rzZo1euONNzRr1izNnz/f5ux2aStovq+7775bGzdu1OjRo9W6dWtVq1ZNeXl56tGjR4E/suHm5laoZZLy/TiMPYX5EZCieOeddyRJI0eO1MiRI/Ot/+CDDzR48OAi7fPCWLz99tsKCgrKt74kfw0cAPB/yLF/cVaOLU25ublF3uadd97RoEGD1Lt3b40ePVq1a9eWm5ub4uPj9eOPP15RPIUdyyt9zACgLCMP/6Ui5uHc3FzdcsstOnHihJ588kk1adJEPj4++uWXXzRo0KACx6ywGjVqJHd3d+3atatI2zkrpxb0fMrLy9Mtt9yiMWPGFLjNNddcY3Of9wPlF9UdoJQFBARo8ODBGjx4sE6fPq0bb7xREydOtN5Y2EumISEh+vLLL3Xq1CmbM/Tff/+9tf7Cv3l5eTp48KAaN25stbv4yuvLOXnypJKSkjRp0iSNHz/eWl6cr+YVx4U+HDhwwDo7Lv31gygZGRlWX4vCGKOlS5fq5ptvzvf1P0maMmWKlixZYhXRQ0JCChyzS5c1bNhQklS7du3LXn0HAChZ5NjLK4kcW9TjF9TXffv25VtWo0YNZWRk2Cw7d+6cjh07ZrOsYcOG+u677/72uO+//76uvvpqffjhhzbPgwkTJti0K0pRw9ljCQBlDXn48pyRO0JCQgrMs5eOr73HZ9euXdq/f78WL16sgQMHWssTExOvOLaqVauqW7duWrt2rY4cOaJ69epd8T6LoijvS+xp2LChTp8+TT2gEmBOdKAUXfrVrmrVqqlRo0bKzs62lvn4+EhSvg+Nt912m3Jzc/XKK6/YLJ81a5ZcXFzUs2dPSbLm53r11Vdt2r388suFjvPCmdFLz4Re+qvSJeW2224r8HgzZ86UpL/99XV7vv76ax06dEiDBw/WnXfeme92zz33aN26dTp69Kikv8YxOTlZO3futPZx4sSJfHOnR0ZGytfXV9OmTSvwV+t//fXXIscKACg6cmzhlESOLerxN23apC1btljLfv3113z5VfrrQ+ml84guWLAg35Xo/fr10zfffKMVK1bk28eFcS5o3Ddv3qzk5GSb9lWrVpWU/zliry+S88YSAMoS8nDhOCN33HbbbdqyZYtNzjtz5owWLFigBg0aKCwsTJL9x6egMTPG6KWXXnJIfBMmTJAxRgMGDNDp06fzrU9JSdHixYsdcqxLFeV9iT133323kpOT9cUXX+Rbl5GRofPnzzskVjgfV6IDpSgsLEw33XST2rZtq4CAAG3btk3vv/++4uLirDZt27aVJD322GOKjIyUm5ub+vfvrzvuuEM333yznnrqKR06dEitWrXSmjVr9PHHH2vEiBHWFdFt27ZVv379NHv2bP3+++/q2LGj1q9fr/3790sq3BVWvr6+uvHGGzVjxgzl5OToH//4h9asWaODBw+WwKjk16pVK8XExGjBggXKyMhQ165dtWXLFi1evFi9e/fWzTffXOR9LlmyRG5ubnbflPzzn//UU089pWXLlmnUqFEaM2aM3nnnHd1yyy169NFH5ePjozfeeEP169fXiRMnrHH09fXVvHnzNGDAAF133XXq37+/rrrqKqWmpmrVqlXq3LlzvjeDAADHI8cWTknk2AvOnz9vTZ12qT59+sjHx0djxozR22+/rR49emj48OHy8fHRggULFBISom+//dZmm/vvv18PP/yw+vXrp1tuuUXffPONvvjii3xz0I4ePVrvv/++7rrrLg0ZMkRt27bViRMn9Mknn2j+/Plq1aqVbr/9dn344Yfq06ePoqKidPDgQc2fP19hYWE2H9i9vb0VFhamd999V9dcc40CAgLUvHnzAueKLcmxBIDyhjxcOM7IHWPHjtV//vMf9ezZU4899pgCAgK0ePFiHTx4UB988IFcXf+6vrZhw4by9/fX/PnzVb16dfn4+KhDhw5q0qSJGjZsqCeeeEK//PKLfH199cEHHzhsDu9OnTpp7ty5euSRR9SkSRMNGDBAjRs31qlTp/TVV1/pk08+0dSpUx1yrEsV5X2JPaNHj9Ynn3yi22+/XYMGDVLbtm115swZ7dq1S++//74OHTqU770LyikDoFASEhKMJLN169YC13ft2tU0a9bMZllISIiJiYmx7k+dOtW0b9/e+Pv7G29vb9OkSRPz7LPPmnPnzlltzp8/bx599FFz1VVXGRcXF3Pxy/TUqVNm5MiRJjg42FSpUsU0btzYPP/88yYvL8/muGfOnDGxsbEmICDAVKtWzfTu3dvs27fPSDLTp0+32k2YMMFIMr/++mu+/vz888+mT58+xt/f3/j5+Zm77rrLHD161EgyEyZMuOw+YmJijI+PT6HGqSA5OTlm0qRJJjQ01FSpUsXUq1fPjBs3zpw9e7ZQx7nYuXPnTM2aNc0NN9zwt+1CQ0NNmzZtrPs7duwwN9xwg/H09DR169Y18fHxZs6cOUaSSUtLs9l23bp1JjIy0vj5+RkvLy/TsGFDM2jQILNt27bL9hUAKjtybPnNsRe3lWT3dvDgQavtt99+a7p27Wq8vLzMP/7xDzNlyhSzcOHCfO1yc3PNk08+aWrVqmWqVq1qIiMjzQ8//JDvsTfGmN9//93ExcWZf/zjH8bDw8PUrVvXxMTEmN9++80YY0xeXp6ZNm2aCQkJMZ6enqZNmzZm5cqVJiYmxoSEhNjsa+PGjaZt27bGw8PD5jG58HgUZyxDQkJMVFRUvnHr2rWr6dq1a6HGGABKCnm4/OfhizVr1ixfbvnxxx/NnXfeafz9/Y2Xl5dp3769WblyZb5tP/74YxMWFmbc3d2NJJOQkGCMMWbPnj0mIiLCVKtWzdSqVcs88MAD5ptvvrFpY0zBubKwUlJSzL/+9S/rOVCjRg3TvXt3s3jxYpObm2uMMebgwYNGknn++efzbW/v8btYQe8hCvu+xF4uN+av5++4ceNMo0aNjIeHh6lVq5bp1KmTeeGFF6zXQFFiR9nkYgwz1wOVwc6dO9WmTRu98847io6OdnY45daIESP02muv6fTp03Z/EAQAULmQYwEAcB7yMIDSwJzoQAX0559/5ls2e/Zsubq66sYbb3RCROXTpeP4+++/6+2331aXLl0ooANAJUWOBQDAecjDAJyFOdGBCmjGjBlKSUnRzTffLHd3d33++ef6/PPP9eCDD5b6r12XZ+Hh4brpppvUtGlTpaena+HChcrKytIzzzzj7NAAAE5CjgUAwHnIw+VHZmZmgSc9LhYUFFRK0QBXjulcgAooMTFRkyZN0p49e3T69GnVr19fAwYM0FNPPSV3d86dFda///1vvf/++/r555/l4uKi6667ThMmTFBERISzQwMAOAk5FgAA5yEPlx+DBg3S4sWL/7YNJUmUJxTRAQAAAAAAADjMnj17dPTo0b9twwVqKE8oogMAAAAAAAAAYAc/LAoAAAAAAAAAgB1MGCUpLy9PR48eVfXq1eXi4uLscAAAlYQxRqdOnVJwcLBcXTmvfSXI5QAAZyCXOwZ5HADgLIXN5RTRJR09epRfcQYAOM2RI0dUt25dZ4dRrpHLAQDORC6/MuRxAICzXS6XU0SXVL16dUl/DZavr6+TowEAVBZZWVmqV6+elYdQfORyAIAzkMsdgzwOAHCWwuZyiuiS9XUxX19fEjYAoNTxteUrRy4HADgTufzKkMcBAM52uVzOpG0AAAAAAAAAANhBER0AAAAAAAAAADsoogMAAAAAAAAAYAdFdAAAAAAAAAAA7HBqEb1BgwZycXHJd4uNjZUknT17VrGxsapZs6aqVaumfv36KT093WYfqampioqKUtWqVVW7dm2NHj1a58+fd0Z3AAAAAAAAAAAVjFOL6Fu3btWxY8esW2JioiTprrvukiSNHDlSn376qZYvX67169fr6NGj6tu3r7V9bm6uoqKidO7cOW3cuFGLFy/WokWLNH78eKf0BwAAAAAAAABQsTi1iH7VVVcpKCjIuq1cuVINGzZU165dlZmZqYULF2rmzJnq1q2b2rZtq4SEBG3cuFGbNm2SJK1Zs0Z79uzRO++8o9atW6tnz56aMmWK5s6dq3PnzjmzawAAAAAAAACACqDMzIl+7tw5vfPOOxoyZIhcXFyUkpKinJwcRUREWG2aNGmi+vXrKzk5WZKUnJysFi1aKDAw0GoTGRmprKws7d692+6xsrOzlZWVZXMDAAAAAAAAAOBSZaaI/tFHHykjI0ODBg2SJKWlpcnDw0P+/v427QIDA5WWlma1ubiAfmH9hXX2xMfHy8/Pz7rVq1fPcR0BAAAAAAAAAFQYZaaIvnDhQvXs2VPBwcElfqxx48YpMzPTuh05cqTEjwkAAAAAAAAAKH/cnR2AJB0+fFhffvmlPvzwQ2tZUFCQzp07p4yMDJur0dPT0xUUFGS12bJli82+0tPTrXX2eHp6ytPT04E9AAAAAAAAAABURGXiSvSEhATVrl1bUVFR1rK2bduqSpUqSkpKspbt27dPqampCg8PlySFh4dr165dOn78uNUmMTFRvr6+CgsLK70OAAAAAAAAAAAqJKdfiZ6Xl6eEhATFxMTI3f3/wvHz89PQoUM1atQoBQQEyNfXV48++qjCw8PVsWNHSdKtt96qsLAwDRgwQDNmzFBaWpqefvppxcbGcqU5AAAAAAAAAOCKOb2I/uWXXyo1NVVDhgzJt27WrFlydXVVv379lJ2drcjISL366qvWejc3N61cuVLDhg1TeHi4fHx8FBMTo8mTJ5dmFwAAAAAAAAAAFZSLMcY4Owhny8rKkp+fnzIzM+Xr6+vscAAAlQT5x3EYSwCAM5B/HINxBAA4S2FzUJmYEx0AAAAAAAAAgLKIIjoAAAAAAAAAAHY4fU50VG4Nxq5ydggFOjQ9ytkhAACcpKzmprKKnAkAAMoK3sdVTrwfRWngSnQAAAAAAAAAAOygiA4AAAAAAAAAgB0U0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADoroAAAAAAAAAADYQREdAAAAAAAAAAA7KKIDAAAAAAAAAGAHRXQAAAAAAAAAAOygiA4AAAAAAAAAgB0U0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADoroAAAAAAAAAADYQREdAAAAAAAAAAA7KKIDAAAAAAAAAGAHRXQAAAAAAAAAAOygiA4AAAAAAAAAgB0U0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADoroAAAAAAAAAADYQREdAAAAAAAAAAA7KKIDAAAAAAAAAGAHRXQAAAAAAAAAAOygiA4AAAAAAAAAgB0U0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADoroAAAAAAAAAADYQREdAAAAAAAAAAA7KKIDAAAAAAAAAGAHRXQAAAAAAAAAAOygiA4AAAAAAAAAgB0U0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADoroAACgQBs2bNAdd9yh4OBgubi46KOPPrJZb4zR+PHjVadOHXl7eysiIkIHDhywaXPixAlFR0fL19dX/v7+Gjp0qE6fPm3T5ttvv9UNN9wgLy8v1atXTzNmzCjprgEAAAAAUGgU0QEAQIHOnDmjVq1aae7cuQWunzFjhubMmaP58+dr8+bN8vHxUWRkpM6ePWu1iY6O1u7du5WYmKiVK1dqw4YNevDBB631WVlZuvXWWxUSEqKUlBQ9//zzmjhxohYsWFDi/QMAoKLjhDgAAI5BER0AABSoZ8+emjp1qvr06ZNvnTFGs2fP1tNPP61evXqpZcuWeuutt3T06FHrA/revXu1evVqvfHGG+rQoYO6dOmil19+WcuWLdPRo0clSUuWLNG5c+f05ptvqlmzZurfv78ee+wxzZw5szS7CgBAhcQJcQAAHIMiOgAAKLKDBw8qLS1NERER1jI/Pz916NBBycnJkqTk5GT5+/urXbt2VpuIiAi5urpq8+bNVpsbb7xRHh4eVpvIyEjt27dPJ0+eLKXeAABQMXFCHAAAx6CIDgAAiiwtLU2SFBgYaLM8MDDQWpeWlqbatWvbrHd3d1dAQIBNm4L2cfExCpKdna2srCybGwAAKDxnnhAnjwMAyhunF9F/+eUX3XfffapZs6a8vb3VokULbdu2zVrvqDnaAABAxREfHy8/Pz/rVq9ePWeHBABAueLME+LkcQBAeePUIvrJkyfVuXNnValSRZ9//rn27NmjF198UTVq1LDaOGKONgAA4FhBQUGSpPT0dJvl6enp1rqgoCAdP37cZv358+d14sQJmzYF7ePiYxRk3LhxyszMtG5Hjhy5sg4BAIBSQx4HAJQ3Ti2iP/fcc6pXr54SEhLUvn17hYaG6tZbb1XDhg0lOW6ONgAA4FihoaEKCgpSUlKStSwrK0ubN29WeHi4JCk8PFwZGRlKSUmx2qxdu1Z5eXnq0KGD1WbDhg3Kycmx2iQmJuraa6+1Oal+KU9PT/n6+trcAABA4TnzhDh5HABQ3ji1iP7JJ5+oXbt2uuuuu1S7dm21adNGr7/+urXeUXO0XYr51wAAuLzTp09r586d2rlzp6S/8vLOnTuVmpoqFxcXjRgxQlOnTtUnn3yiXbt2aeDAgQoODlbv3r0lSU2bNlWPHj30wAMPaMuWLfr6668VFxen/v37Kzg4WJL0r3/9Sx4eHho6dKh2796td999Vy+99JJGjRrlpF4DAFA5OPuEOAAA5YlTi+g//fST5s2bp8aNG+uLL77QsGHD9Nhjj2nx4sWSHDdH26WYfw0AgMvbtm2b2rRpozZt2kiSRo0apTZt2mj8+PGSpDFjxujRRx/Vgw8+qOuvv16nT5/W6tWr5eXlZe1jyZIlatKkibp3767bbrtNXbp00YIFC6z1fn5+WrNmjQ4ePKi2bdvq8ccf1/jx45mWDQAAB+CEOAAAjuHuzIPn5eWpXbt2mjZtmiSpTZs2+u677zR//nzFxMSU2HHHjRtnk9CzsrIopAMAcImbbrpJxhi7611cXDR58mRNnjzZbpuAgAAtXbr0b4/TsmVL/fe//y12nAAAoGDbtm3TzTffbN2/8Dk4JiZGixYt0pgxY3TmzBk9+OCDysjIUJcuXQo8IR4XF6fu3bvL1dVV/fr105w5c6z1F06Ix8bGqm3btqpVqxYnxAEAFY5Ti+h16tRRWFiYzbKmTZvqgw8+kGQ7R1udOnWsNunp6WrdurXV5nJztF3K09NTnp6ejuoGAAAAAABlDifEAQBwDKdO59K5c2ft27fPZtn+/fsVEhIiyXFztAEAAAAAAAAAUBxOvRJ95MiR6tSpk6ZNm6a7775bW7Zs0YIFC6y5Ui+eo61x48YKDQ3VM888Y3eOtvnz5ysnJyffHG0AAAAAAAAAABSHU4vo119/vVasWKFx48Zp8uTJCg0N1ezZsxUdHW21ccQcbQAAAAAAAAAAFIdTi+iSdPvtt+v222+3u95Rc7QBAAAAAAAAAFBUTp0THQAAAAAAAACAsowiOgAAAAAAAAAAdlBEBwAAAAAAAADADoroAAAAAAAAAADYQREdAAAAAAAAAAA7KKIDAAAAAAAAAGAHRXQAAAAAAAAAAOygiA4AAAAAAAAAgB0U0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADoroAAAAAAAAAADYQREdAAAAAAAAAAA7KKIDAAAAAAAAAGAHRXQAAAAAAAAAAOygiA4AAAAAAAAAgB0U0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADoroAAAAAAAAAADYQREdAAAAAAAAAAA7KKIDAAAAAAAAAGAHRXQAAAAAAAAAAOygiA4AAAAAAAAAgB0U0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADndnBwCURQ3GrnJ2CAU6ND3K2SEAAAAAAAAAlQpXogMAAAAAAAAAYAdFdAAAAAAAAAAA7KCIDgAAAAAAAACAHRTRAQAAAAAAAACwgyI6AAAAAAAAAAB2UEQHAAAAAAAAAMAOiugAAAAAAAAAANhBER0AAAAAAAAAADsoogMAAAAAAAAAYAdFdAAAAAAAAAAA7KCIDgAAAAAAAACAHRTRAQAAAAAAAACwgyI6AAAAAAAAAAB2UEQHAAAAAAAAAMAOiugAAAAAAAAAANhBER0AAAAAAAAAADsoogMAAAAAAAAAYIdTi+gTJ06Ui4uLza1JkybW+rNnzyo2NlY1a9ZUtWrV1K9fP6Wnp9vsIzU1VVFRUapatapq166t0aNH6/z586XdFQAAAAAAAABABeTu7ACaNWumL7/80rrv7v5/IY0cOVKrVq3S8uXL5efnp7i4OPXt21dff/21JCk3N1dRUVEKCgrSxo0bdezYMQ0cOFBVqlTRtGnTSr0vZVmDsaucHQIAAAAAAAAAlDtOL6K7u7srKCgo3/LMzEwtXLhQS5cuVbdu3SRJCQkJatq0qTZt2qSOHTtqzZo12rNnj7788ksFBgaqdevWmjJlip588klNnDhRHh4epd0dAAAAAAAAAEAF4vQ50Q8cOKDg4GBdffXVio6OVmpqqiQpJSVFOTk5ioiIsNo2adJE9evXV3JysiQpOTlZLVq0UGBgoNUmMjJSWVlZ2r17d+l2BAAAAAAAAABQ4Tj1SvQOHTpo0aJFuvbaa3Xs2DFNmjRJN9xwg7777julpaXJw8ND/v7+NtsEBgYqLS1NkpSWlmZTQL+w/sI6e7Kzs5WdnW3dz8rKclCPAAAAAAAAAAAViVOL6D179rT+37JlS3Xo0EEhISF677335O3tXWLHjY+P16RJk0ps/wAAAAAAAACAisHp07lczN/fX9dcc41++OEHBQUF6dy5c8rIyLBpk56ebs2hHhQUpPT09HzrL6yzZ9y4ccrMzLRuR44ccWxHAAAAAAAAAAAVQpkqop8+fVo//vij6tSpo7Zt26pKlSpKSkqy1u/bt0+pqakKDw+XJIWHh2vXrl06fvy41SYxMVG+vr4KCwuzexxPT0/5+vra3AAAAAAAAAAAuJRTi+hPPPGE1q9fr0OHDmnjxo3q06eP3NzcdO+998rPz09Dhw7VqFGjtG7dOqWkpGjw4MEKDw9Xx44dJUm33nqrwsLCNGDAAH3zzTf64osv9PTTTys2Nlaenp7O7BoAABVebm6unnnmGYWGhsrb21sNGzbUlClTZIyx2hhjNH78eNWpU0fe3t6KiIjQgQMHbPZz4sQJRUdHy9fXV/7+/ho6dKhOnz5d2t0BAKDSIZcDAFA4Ti2i//zzz7r33nt17bXX6u6771bNmjW1adMmXXXVVZKkWbNm6fbbb1e/fv104403KigoSB9++KG1vZubm1auXCk3NzeFh4frvvvu08CBAzV58mRndQkAgErjueee07x58/TKK69o7969eu655zRjxgy9/PLLVpsZM2Zozpw5mj9/vjZv3iwfHx9FRkbq7NmzVpvo6Gjt3r1biYmJWrlypTZs2KAHH3zQGV0CAKBSIZcDAFA4LubiU8yVVFZWlvz8/JSZmVlhp3ZpMHaVs0OAAxyaHuXsEAA4UHnPP7fffrsCAwO1cOFCa1m/fv3k7e2td955R8YYBQcH6/HHH9cTTzwhScrMzFRgYKAWLVqk/v37a+/evQoLC9PWrVvVrl07SdLq1at122236eeff1ZwcHChYnHkWJIzi4bcBKAyI5c7JpeX93FE2cH7uMqJ96O4EoXNQWVqTnQAAFB+dOrUSUlJSdq/f78k6ZtvvtH//vc/9ezZU5J08OBBpaWlKSIiwtrGz89PHTp0UHJysiQpOTlZ/v7+1oduSYqIiJCrq6s2b95s99jZ2dnKysqyuQEAgKJxZi4HAKA8cXd2AAAAoHwaO3assrKy1KRJE7m5uSk3N1fPPvusoqOjJUlpaWmSpMDAQJvtAgMDrXVpaWmqXbu2zXp3d3cFBARYbQoSHx+vSZMmObI7AABUOs7K5dnZ2crOzrbuczIcAFDWcSU6AAAolvfee09LlizR0qVLtX37di1evFgvvPCCFi9eXOLHHjdunDIzM63bkSNHSvyYAABUNM7K5fHx8fLz87Nu9erVK9HjAQBwpbgSHQAAFMvo0aM1duxY9e/fX5LUokULHT58WPHx8YqJiVFQUJAkKT09XXXq1LG2S09PV+vWrSVJQUFBOn78uM1+z58/rxMnTljbF8TT01Oenp4O7hEAAJWLs3L5uHHjNGrUKOt+VlYWhXQAQJnGlegAAKBY/vjjD7m62r6VcHNzU15eniQpNDRUQUFBSkpKstZnZWVp8+bNCg8PlySFh4crIyNDKSkpVpu1a9cqLy9PHTp0KIVeAABQeTkrl3t6esrX19fmBgBAWcaV6AAAoFjuuOMOPfvss6pfv76aNWumHTt2aObMmRoyZIgkycXFRSNGjNDUqVPVuHFjhYaG6plnnlFwcLB69+4tSWratKl69OihBx54QPPnz1dOTo7i4uLUv39/BQcHO7F3AABUfORyAAAKhyI6AAAolpdfflnPPPOMHnnkER0/flzBwcF66KGHNH78eKvNmDFjdObMGT344IPKyMhQly5dtHr1anl5eVltlixZori4OHXv3l2urq7q16+f5syZ44wuAQBQqZDLAQAoHBdjjHF2EM6WlZUlPz8/ZWZmVtivkTUYu8rZIcABDk2PcnYIAByoMuSf0uLIsSRnFg25CUBlRi53DMYRjsL7uMqJ96O4EoXNQcyJDgAAAAAAAACAHRTRAQAAAAAAAACwgyI6AAAAAAAAAAB2UEQHAAAAAAAAAMAOiugAAAAAAAAAANhBER0AAAAAAAAAADsoogMAAAAAAAAAYAdFdAAAAAAAAAAA7KCIDgAAAAAAAACAHRTRAQAAAAAAAACwgyI6AAAAAAAAAAB2UEQHAAAAAAAAAMAOiugAAAAAAAAAANhBER0AAAAAAAAAADsoogMAAAAAAAAAYAdFdAAAAAAAAAAA7KCIDgAAAAAAAACAHRTRAQAAAAAAAACwgyI6AAAAAAAAAAB2UEQHAAAAAAAAAMAOiugAAAAAAAAAANhBER0AAAAAAAAAADsoogMAAAAAAAAAYAdFdAAAAAAAAAAA7KCIDgAAAAAAAACAHRTRAQAAAAAAAACwgyI6AAAAAAAAAAB2UEQHAAAAAAAAAMAOiugAAAAAAAAAANhBER0AAAAAAAAAADsoogMAAAAAAAAAYEexiug//fSTo+MAAAAORK4GAKD8Io8DAFC2FKuI3qhRI91888165513dPbsWUfHBAAArhC5GgCA8os8DgBA2VKsIvr27dvVsmVLjRo1SkFBQXrooYe0ZcsWR8cGAACKiVwNAED5RR4HAKBsKVYRvXXr1nrppZd09OhRvfnmmzp27Ji6dOmi5s2ba+bMmfr1118dHScAACgCcjUAAOUXeRwAgLLlin5Y1N3dXX379tXy5cv13HPP6YcfftATTzyhevXqaeDAgTp27Jij4gQAAMVArgYAoPwijwMAUDZcURF927ZteuSRR1SnTh3NnDlTTzzxhH788UclJibq6NGj6tWrl6PiBAAAxUCuBgCg/CKPAwBQNhSriD5z5ky1aNFCnTp10tGjR/XWW2/p8OHDmjp1qkJDQ3XDDTdo0aJF2r59e6H3OX36dLm4uGjEiBHWsrNnzyo2NlY1a9ZUtWrV1K9fP6Wnp9tsl5qaqqioKFWtWlW1a9fW6NGjdf78+eJ0CwCACqMkcjUAACgd5HEAAMoW9+JsNG/ePA0ZMkSDBg1SnTp1CmxTu3ZtLVy4sFD727p1q1577TW1bNnSZvnIkSO1atUqLV++XH5+foqLi1Pfvn319ddfS5Jyc3MVFRWloKAgbdy4UceOHdPAgQNVpUoVTZs2rThdAwCgQnB0rgYAAKWHPA4AQNlSrCL6gQMHLtvGw8NDMTExl213+vRpRUdH6/XXX9fUqVOt5ZmZmVq4cKGWLl2qbt26SZISEhLUtGlTbdq0SR07dtSaNWu0Z88effnllwoMDFTr1q01ZcoUPfnkk5o4caI8PDyK0z0AAMo9R+ZqAABQusjjAACULcWaziUhIUHLly/Pt3z58uVavHhxkfYVGxurqKgoRURE2CxPSUlRTk6OzfImTZqofv36Sk5OliQlJyerRYsWCgwMtNpERkYqKytLu3fvLlIcAABUJI7M1QAAoHSRxwEAKFuKVUSPj49XrVq18i2vXbt2kaZRWbZsmbZv3674+Ph869LS0uTh4SF/f3+b5YGBgUpLS7PaXFxAv7D+wjp7srOzlZWVZXMDAKAicVSuBgAApY88DgBA2VKsInpqaqpCQ0PzLQ8JCVFqamqh9nHkyBENHz5cS5YskZeXV3HCKLb4+Hj5+flZt3r16pXq8QEAKGmOyNUAAMA5yOMAAJQtxSqi165dW99++22+5d98841q1qxZqH2kpKTo+PHjuu666+Tu7i53d3etX79ec+bMkbu7uwIDA3Xu3DllZGTYbJeenq6goCBJUlBQkNLT0/Otv7DOnnHjxikzM9O6HTlypFAxAwBQXjgiVwMAAOcgjwMAULYUq4h+77336rHHHtO6deuUm5ur3NxcrV27VsOHD1f//v0LtY/u3btr165d2rlzp3Vr166doqOjrf9XqVJFSUlJ1jb79u1TamqqwsPDJUnh4eHatWuXjh8/brVJTEyUr6+vwsLC7B7b09NTvr6+NjcAACoSR+RqAADgHORxAADKFvfibDRlyhQdOnRI3bt3l7v7X7vIy8vTwIEDCz0/W/Xq1dW8eXObZT4+PqpZs6a1fOjQoRo1apQCAgLk6+urRx99VOHh4erYsaMk6dZbb1VYWJgGDBigGTNmKC0tTU8//bRiY2Pl6elZnK4BAFAhOCJXAwAA5yCPAwBQthSriO7h4aF3331XU6ZM0TfffCNvb2+1aNFCISEhDg1u1qxZcnV1Vb9+/ZSdna3IyEi9+uqr1no3NzetXLlSw4YNU3h4uHx8fBQTE6PJkyc7NA4AAMqb0srVAADA8cjjAACULcUqol9wzTXX6JprrnFULPrqq69s7nt5eWnu3LmaO3eu3W1CQkL02WefOSwGAAAqEkfnagAAUHrI4wAAlA3FKqLn5uZq0aJFSkpK0vHjx5WXl2ezfu3atQ4JDgAAFA+5GgCA8os8DgBA2VKsIvrw4cO1aNEiRUVFqXnz5nJxcXF0XAAA4AqQqwEAKL/I41emwdhVzg4BAFDBFKuIvmzZMr333nu67bbbHB0PAABwAHI1AADlF3kcAICyxbU4G3l4eKhRo0aOjgUAADgIuRoAgPKLPA4AQNlSrCL6448/rpdeeknGGEfHAwAAHIBcDQBA+UUeBwCgbCnWdC7/+9//tG7dOn3++edq1qyZqlSpYrP+ww8/dEhwAACgeMjVAACUX+RxAADKlmIV0f39/dWnTx9HxwIAAByEXA0AQPlFHgcAoGwpVhE9ISHB0XEAAAAHIlcDAFB+kccBAChbijUnuiSdP39eX375pV577TWdOnVKknT06FGdPn3aYcEBAIDiI1cDAFB+kccBACg7ilVEP3z4sFq0aKFevXopNjZWv/76qyTpueee0xNPPOHQAAEAQNGVVq7+5ZdfdN9996lmzZry9vZWixYttG3bNmu9MUbjx49XnTp15O3trYiICB04cMBmHydOnFB0dLR8fX3l7++voUOHUiAAAFRqpfmZm1wOAMDlFauIPnz4cLVr104nT56Ut7e3tbxPnz5KSkpyWHAAAKB4SiNXnzx5Up07d1aVKlX0+eefa8+ePXrxxRdVo0YNq82MGTM0Z84czZ8/X5s3b5aPj48iIyN19uxZq010dLR2796txMRErVy5Uhs2bNCDDz7okBgBACiPSuszN7kcAIDCKdac6P/973+1ceNGeXh42Cxv0KCBfvnlF4cEBgAAiq80cvVzzz2nevXq2czbGhoaav3fGKPZs2fr6aefVq9evSRJb731lgIDA/XRRx+pf//+2rt3r1avXq2tW7eqXbt2kqSXX35Zt912m1544QUFBwc7JFYAAMqT0vrMTS4HAKBwinUlel5ennJzc/Mt//nnn1W9evUrDgoAAFyZ0sjVn3zyidq1a6e77rpLtWvXVps2bfT6669b6w8ePKi0tDRFRERYy/z8/NShQwclJydLkpKTk+Xv72996JakiIgIubq6avPmzQ6JEwCA8qa0PnOTywEAKJxiFdFvvfVWzZ4927rv4uKi06dPa8KECbrtttscFRsAACim0sjVP/30k+bNm6fGjRvriy++0LBhw/TYY49p8eLFkqS0tDRJUmBgoM12gYGB1rq0tDTVrl3bZr27u7sCAgKsNgXJzs5WVlaWzQ0AgIqitD5zOyuXk8cBAOVNsaZzefHFFxUZGamwsDCdPXtW//rXv3TgwAHVqlVL//nPfxwdIwAAKKLSyNV5eXlq166dpk2bJklq06aNvvvuO82fP18xMTEOOYY98fHxmjRpUokeAwAAZymtz9zOyuXkcQBAeVOsInrdunX1zTffaNmyZfr22291+vRpDR06VNHR0TY/egIAAJyjNHJ1nTp1FBYWZrOsadOm+uCDDyRJQUFBkqT09HTVqVPHapOenq7WrVtbbY4fP26zj/Pnz+vEiRPW9gUZN26cRo0aZd3PyspSvXr1rqg/AACUFaX1mdtZuZw8DgAob4pVRJf++nrWfffd58hYAACAA5V0ru7cubP27dtns2z//v0KCQmR9NcPkwUFBSkpKcn6oJ2VlaXNmzdr2LBhkqTw8HBlZGQoJSVFbdu2lSStXbtWeXl56tChg91je3p6ytPTswR6BQBA2VAan7mdlcvJ4wCA8qZYRfS33nrrb9cPHDiwWMEAAADHKI1cPXLkSHXq1EnTpk3T3XffrS1btmjBggVasGCBpL/mbx0xYoSmTp2qxo0bKzQ0VM8884yCg4PVu3dvSX9d7dajRw898MADmj9/vnJychQXF6f+/fsrODj4imMEAKA8Kq3P3ORyAAAKp1hF9OHDh9vcz8nJ0R9//CEPDw9VrVqVIjoAAE5WGrn6+uuv14oVKzRu3DhNnjxZoaGhmj17tqKjo602Y8aM0ZkzZ/Tggw8qIyNDXbp00erVq+Xl5WW1WbJkieLi4tS9e3e5urqqX79+mjNnzhXHBwBAeVVan7nJ5QAAFI6LMcY4YkcHDhzQsGHDNHr0aEVGRjpil6UmKytLfn5+yszMlK+vr7PDKRENxq5ydghwgEPTo5wdAgAHKu38U55z9eU4cizJmUVDbgJQmZVmLiePFx65HKhceD+KK1HYHOTqqAM2btxY06dPz3fGHAAAlA3kagAAyi/yOAAAzuOwIrr01w+fHD161JG7BAAADkSuBgCg/CKPAwDgHMWaE/2TTz6xuW+M0bFjx/TKK6+oc+fODgkMAAAUH7kaAIDyizwOAEDZUqwi+oVf4b7AxcVFV111lbp166YXX3zREXEBAIArQK4GAKD8Io8DAFC2FKuInpeX5+g4AACAA5GrAQAov8jjAACULQ6dEx0AAAAAAAAAgIqkWFeijxo1qtBtZ86cWZxDAACAK0CuBgCg/CKPAwBQthSriL5jxw7t2LFDOTk5uvbaayVJ+/fvl5ubm6677jqrnYuLi2OiBAAARUKuBgCg/CKPAwBQthSriH7HHXeoevXqWrx4sWrUqCFJOnnypAYPHqwbbrhBjz/+uEODBAAARUOuBgCg/CKPAwBQthRrTvQXX3xR8fHxVjKXpBo1amjq1Kn8UjgAAGUAuRoAgPKLPA4AQNlSrCJ6VlaWfv3113zLf/31V506deqKgwIAAFeGXA0AQPlFHgcAoGwpVhG9T58+Gjx4sD788EP9/PPP+vnnn/XBBx9o6NCh6tu3r6NjBAAARUSuBgCg/CKPAwBQthRrTvT58+friSee0L/+9S/l5OT8tSN3dw0dOlTPP/+8QwMEAABFR64GAKD8Io8DAFC2FKuIXrVqVb366qt6/vnn9eOPP0qSGjZsKB8fH4cGBwAAiodcDQBA+UUeBwCgbCnWdC4XHDt2TMeOHVPjxo3l4+MjY4yj4gIAAA5ArgYAoPwijwMAUDYUq4j++++/q3v37rrmmmt022236dixY5KkoUOH6vHHH3dogAAAoOjI1QAAlF/kcQAAypZiFdFHjhypKlWqKDU1VVWrVrWW33PPPVq9erXDggMAAMVDrgYAoPwijwMAULYUa070NWvW6IsvvlDdunVtljdu3FiHDx92SGAAAKD4yNUAAJRf5HEAAMqWYl2JfubMGZuz4RecOHFCnp6eVxwUAAC4MuRqAADKL/I4AABlS7GK6DfccIPeeust676Li4vy8vI0Y8YM3XzzzQ4LDgAAFA+5GgCA8os8DgBA2VKs6VxmzJih7t27a9u2bTp37pzGjBmj3bt368SJE/r6668dHSMAACgicjUAAOUXeRwAgLKlWFeiN2/eXPv371eXLl3Uq1cvnTlzRn379tWOHTvUsGFDR8cIAACKiFwNAED5RR4HAKBsKfKV6Dk5OerRo4fmz5+vp556qiRiAgAAV4BcDQBA+UUeBwCg7CnylehVqlTRt99+WxKxAAAAByBXAwBQfpHHAQAoe4o1nct9992nhQsXOjoWAADgIORqAADKL/I4AABlS7F+WPT8+fN688039eWXX6pt27by8fGxWT9z5sxC7WfevHmaN2+eDh06JElq1qyZxo8fr549e0qSzp49q8cff1zLli1Tdna2IiMj9eqrryowMNDaR2pqqoYNG6Z169apWrVqiomJUXx8vNzdi9U1AAAqBEflagAAUPrI4wAAlC1FqjT/9NNPatCggb777jtdd911kqT9+/fbtHFxcSn0/urWravp06ercePGMsZo8eLF6tWrl3bs2KFmzZpp5MiRWrVqlZYvXy4/Pz/FxcWpb9++1q+R5+bmKioqSkFBQdq4caOOHTumgQMHqkqVKpo2bVpRugYAQIXg6FwNAABKD3kcAICyqUhF9MaNG+vYsWNat26dJOmee+7RnDlzbK4ML4o77rjD5v6zzz6refPmadOmTapbt64WLlyopUuXqlu3bpKkhIQENW3aVJs2bVLHjh21Zs0a7dmzR19++aUCAwPVunVrTZkyRU8++aQmTpwoDw+PYsUFAEB55ehcDQAASg95HACAsqlIc6IbY2zuf/755zpz5oxDAsnNzdWyZct05swZhYeHKyUlRTk5OYqIiLDaNGnSRPXr11dycrIkKTk5WS1atLB5QxEZGamsrCzt3r3bIXEBAFCelGSuBgAAJYs8DgBA2XRFE4dfmuCLY9euXQoPD9fZs2dVrVo1rVixQmFhYdq5c6c8PDzk7+9v0z4wMFBpaWmSpLS0tHxn5C/cv9CmINnZ2crOzrbuZ2VlXXE/AAAoixyRqwEAgHOQxwEAKBuKVER3cXHJN//alc7Hdu2112rnzp3KzMzU+++/r5iYGK1fv/6K9nk58fHxmjRpUokeAygJDcaucnYIBTo0PcrZIQD4/0oiVwMAgNJBHgcAoGwqUhHdGKNBgwbJ09NTknT27Fk9/PDD+X4p/MMPPyz0Pj08PNSoUSNJUtu2bbV161a99NJLuueee3Tu3DllZGTYXI2enp6uoKAgSVJQUJC2bNlis7/09HRrnT3jxo3TqFGjrPtZWVmqV69eoWMGAKCsKolcDQAASgd5HACAsqlIRfSYmBib+/fdd59Dg5GkvLw8ZWdnq23btqpSpYqSkpLUr18/SdK+ffuUmpqq8PBwSVJ4eLieffZZHT9+XLVr15YkJSYmytfXV2FhYXaP4enpab0pAQCgIimNXA0AAEoGeRwAgLKpSEX0hIQEhx583Lhx6tmzp+rXr69Tp05p6dKl+uqrr/TFF1/Iz89PQ4cO1ahRoxQQECBfX189+uijCg8PV8eOHSVJt956q8LCwjRgwADNmDFDaWlpevrppxUbG0uRHABQKTk6VwOXU1anGiurmAINwN8hjwMAUDZd0Q+LXqnjx49r4MCBOnbsmPz8/NSyZUt98cUXuuWWWyRJs2bNkqurq/r166fs7GxFRkbq1VdftbZ3c3PTypUrNWzYMIWHh8vHx0cxMTGaPHmys7oEAAAAAAAAAKhAnFpEX7hw4d+u9/Ly0ty5czV37ly7bUJCQvTZZ585OjQAAAAAAAAAAOTq7AAAAAAAAAAAACirnHolOgAAAAAAAAAUF7/RUzmV9m8NcSU6AAAAAAAAAAB2UEQHAAAAAAAAAMAOiugAAAAAAAAAANhBER0AAAAAAAAAADsoogMAAAAAAAAAYAdFdAAAAAAAAAAA7KCIDgAAAAAAAACAHRTRAQAAAAAAAACwgyI6AAAAAAAAAAB2UEQHAAAAAAAAAMAOiugAAAAAAAAAANhBER0AAAAAAAAAADsoogMAAAAAAAAAYAdFdAAAAAAAAAAA7KCIDgAAAAAAAACAHRTRAQAAAAAAAACwgyI6AAAAAAAAAAB2UEQHAAAAAAAAAMAOiugAAAAAAAAAANhBER0AADjE9OnT5eLiohEjRljLzp49q9jYWNWsWVPVqlVTv379lJ6ebrNdamqqoqKiVLVqVdWuXVujR4/W+fPnSzl6AAAAAAAKRhEdAABcsa1bt+q1115Ty5YtbZaPHDlSn376qZYvX67169fr6NGj6tu3r7U+NzdXUVFROnfunDZu3KjFixdr0aJFGj9+fGl3AQCASo8T4gAAFIwiOgAAuCKnT59WdHS0Xn/9ddWoUcNanpmZqYULF2rmzJnq1q2b2rZtq4SEBG3cuFGbNm2SJK1Zs0Z79uzRO++8o9atW6tnz56aMmWK5s6dq3PnzjmrSwAAVDqcEAcAwD6K6AAA4IrExsYqKipKERERNstTUlKUk5Njs7xJkyaqX7++kpOTJUnJyclq0aKFAgMDrTaRkZHKysrS7t27S6cDAABUcpwQBwDg71FEBwAAxbZs2TJt375d8fHx+dalpaXJw8ND/v7+NssDAwOVlpZmtbm4gH5h/YV19mRnZysrK8vmBgAAiqe0T4iTxwEA5Q1FdAAAUCxHjhzR8OHDtWTJEnl5eZXqsePj4+Xn52fd6tWrV6rHBwCgonDGCXHyOACgvKGIDgAAiiUlJUXHjx/XddddJ3d3d7m7u2v9+vWaM2eO3N3dFRgYqHPnzikjI8Nmu/T0dAUFBUmSgoKC8v042YX7F9oUZNy4ccrMzLRuR44ccWznAACoBJx1Qpw8DgAobyiiAwCAYunevbt27dqlnTt3Wrd27dopOjra+n+VKlWUlJRkbbNv3z6lpqYqPDxckhQeHq5du3bp+PHjVpvExET5+voqLCzM7rE9PT3l6+trcwMAAEXjrBPi5HEAQHnj7uwAAABA+VS9enU1b97cZpmPj49q1qxpLR86dKhGjRqlgIAA+fr66tFHH1V4eLg6duwoSbr11lsVFhamAQMGaMaMGUpLS9PTTz+t2NhYeXp6lnqfAACoTC6cEL/Y4MGD1aRJEz355JOqV6+edUK8X79+kgo+If7ss8/q+PHjql27tqTCnRAHAKA8oYgOAABKzKxZs+Tq6qp+/fopOztbkZGRevXVV631bm5uWrlypYYNG6bw8HD5+PgoJiZGkydPdmLUAABUDpwQBwCgcCiiAwAAh/nqq69s7nt5eWnu3LmaO3eu3W1CQkL02WeflXBkAACgODghDgAARXQAAAAAAPD/cUIcAID8+GFRAAAAAAAAAADsoIgOAAAAAAAAAIAdFNEBAAAAAAAAALCDOdEBAACAUtJg7Cpnh1CuHJoe5ewQAAAAAK5EBwAAAAAAAADAHoroAAAAAAAAAADYQREdAAAAAAAAAAA7KKIDAAAAAAAAAGAHRXQAAAAAAAAAAOygiA4AAAAAAAAAgB0U0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADqcW0ePj43X99derevXqql27tnr37q19+/bZtDl79qxiY2NVs2ZNVatWTf369VN6erpNm9TUVEVFRalq1aqqXbu2Ro8erfPnz5dmVwAAAAAAAAAAFZBTi+jr169XbGysNm3apMTEROXk5OjWW2/VmTNnrDYjR47Up59+quXLl2v9+vU6evSo+vbta63Pzc1VVFSUzp07p40bN2rx4sVatGiRxo8f74wuAQAAAAAAAAAqEHdnHnz16tU29xctWqTatWsrJSVFN954ozIzM7Vw4UItXbpU3bp1kyQlJCSoadOm2rRpkzp27Kg1a9Zoz549+vLLLxUYGKjWrVtrypQpevLJJzVx4kR5eHg4o2sAAAAAAAAAgAqgTM2JnpmZKUkKCAiQJKWkpCgnJ0cRERFWmyZNmqh+/fpKTk6WJCUnJ6tFixYKDAy02kRGRiorK0u7d+8u8DjZ2dnKysqyuQEAAAAAAAAAcKkyU0TPy8vTiBEj1LlzZzVv3lySlJaWJg8PD/n7+9u0DQwMVFpamtXm4gL6hfUX1hUkPj5efn5+1q1evXoO7g0AAAAAAAAAoCIoM0X02NhYfffdd1q2bFmJH2vcuHHKzMy0bkeOHCnxYwIAAAAAAAAAyh+nzol+QVxcnFauXKkNGzaobt261vKgoCCdO3dOGRkZNlejp6enKygoyGqzZcsWm/2lp6db6wri6ekpT09PB/cCAAAAAAAAAFDROPVKdGOM4uLitGLFCq1du1ahoaE269u2basqVaooKSnJWrZv3z6lpqYqPDxckhQeHq5du3bp+PHjVpvExET5+voqLCysdDoCAAAAAAAAAKiQnHolemxsrJYuXaqPP/5Y1atXt+Yw9/Pzk7e3t/z8/DR06FCNGjVKAQEB8vX11aOPPqrw8HB17NhRknTrrbcqLCxMAwYM0IwZM5SWlqann35asbGxXG0OAAAAAAAAALgiTi2iz5s3T5J000032SxPSEjQoEGDJEmzZs2Sq6ur+vXrp+zsbEVGRurVV1+12rq5uWnlypUaNmyYwsPD5ePjo5iYGE2ePLm0ugEAAAAAAAAAqKCcWkQ3xly2jZeXl+bOnau5c+fabRMSEqLPPvvMkaFdkQZjVzk7BAAAAAAAAACAAzh1TnQAAAAAAAAAAMoyiugAAAAAAAAAANhBER0AAAAAAAAAADsoogMAAAAAAAAAYAdFdAAAAAAAAAAA7KCIDgAAAAAAAACAHRTRAQAAAAAAAACwgyI6AAAAAAAAAAB2UEQHAAAAAAAAAMAOiugAAAAAAAAAANhBER0AAAAAAAAAADsoogMAAAAAAAAAYAdFdAAAAAAAAAAA7KCIDgAAAAAAAACAHRTRAQAAAAAAAACwgyI6AAAAAAAAAAB2UEQHAAAAAAAAAMAOiugAAAAAAAAAANhBER0AAAAAAAAAADsoogMAAAAAAAAAYAdFdAAAAAAAAAAA7KCIDgAAAAAAAACAHRTRAQAAAAAAAACwgyI6AAAAAAAAAAB2UEQHAAAAAAAAAMAOiugAAAAAAAAAANhBER0AAAAAAAAAADsoogMAAAAAAAAAYAdFdAAAAAAAAAAA7KCIDgAAAAAAAACAHRTRAQAAAAAAAACwgyI6AAAAAAAAAAB2UEQHAAAAAAAAAMAOiugAAKDY4uPjdf3116t69eqqXbu2evfurX379tm0OXv2rGJjY1WzZk1Vq1ZN/fr1U3p6uk2b1NRURUVFqWrVqqpdu7ZGjx6t8+fPl2ZXAACodMjjAAAUDkV0AABQbOvXr1dsbKw2bdqkxMRE5eTk6NZbb9WZM2esNiNHjtSnn36q5cuXa/369Tp69Kj69u1rrc/NzVVUVJTOnTunjRs3avHixVq0aJHGjx/vjC4BAFBpkMcBACgcF2OMcXYQzpaVlSU/Pz9lZmbK19f3ivfXYOwqB0QFlB+Hpkc5OwSgXHJ0/ikLfv31V9WuXVvr16/XjTfeqMzMTF111VVaunSp7rzzTknS999/r6ZNmyo5OVkdO3bU559/rttvv11Hjx5VYGCgJGn+/Pl68skn9euvv8rDw+Oyx3XkWJLHgbKD9xgo6ypaLq8IeVwilwNAZeCo94mFzUHuDjkaAACApMzMTElSQECAJCklJUU5OTmKiIiw2jRp0kT169e3PnwnJyerRYsW1gdvSYqMjNSwYcO0e/dutWnTJt9xsrOzlZ2dbd3PysoqqS4BcCIKYYXHCQc4AnkcAICCMZ0LAABwiLy8PI0YMUKdO3dW8+bNJUlpaWny8PCQv7+/TdvAwEClpaVZbS7+4H1h/YV1BYmPj5efn591q1evnoN7AwBA5UIeBwDAPoroAADAIWJjY/Xdd99p2bJlJX6scePGKTMz07odOXKkxI8JAEBFRh4HAMA+pnMBAABXLC4uTitXrtSGDRtUt25da3lQUJDOnTunjIwMm6vY0tPTFRQUZLXZsmWLzf7S09OtdQXx9PSUp6eng3sBAEDlRB4HAODvcSU6AAAoNmOM4uLitGLFCq1du1ahoaE269u2basqVaooKSnJWrZv3z6lpqYqPDxckhQeHq5du3bp+PHjVpvExET5+voqLCysdDoCAEAlRB4HAKBwuBIdAAAUW2xsrJYuXaqPP/5Y1atXt+Y+9fPzk7e3t/z8/DR06FCNGjVKAQEB8vX11aOPPqrw8HB17NhRknTrrbcqLCxMAwYM0IwZM5SWlqann35asbGxXKUGAEAJIo8DAFA4FNEBAECxzZs3T5J000032SxPSEjQoEGDJEmzZs2Sq6ur+vXrp+zsbEVGRurVV1+12rq5uWnlypUaNmyYwsPD5ePjo5iYGE2ePLm0ugEAQKVEHgcAoHAoogMAgGIzxly2jZeXl+bOnau5c+fabRMSEqLPPvvMkaEBAIDLII8DAFA4zIkOAAAAAAAAAIAdTi2ib9iwQXfccYeCg4Pl4uKijz76yGa9MUbjx49XnTp15O3trYiICB04cMCmzYkTJxQdHS1fX1/5+/tr6NChOn36dCn2AgAAAAAAAABQUTm1iH7mzBm1atXK7tfCZsyYoTlz5mj+/PnavHmzfHx8FBkZqbNnz1ptoqOjtXv3biUmJmrlypXasGGDHnzwwdLqAgAAAAAAAACgAnPqnOg9e/ZUz549C1xnjNHs2bP19NNPq1evXpKkt956S4GBgfroo4/Uv39/7d27V6tXr9bWrVvVrl07SdLLL7+s2267TS+88IKCg4NLrS8AAAAAAAAAgIqnzM6JfvDgQaWlpSkiIsJa5ufnpw4dOig5OVmSlJycLH9/f6uALkkRERFydXXV5s2b7e47OztbWVlZNjcAAAAAAAAAAC5VZovoaWlpkqTAwECb5YGBgda6tLQ01a5d22a9u7u7AgICrDYFiY+Pl5+fn3WrV6+eg6MHAAAAAAAAAFQEZbaIXpLGjRunzMxM63bkyBFnhwQAAAAAAAAAKIPKbBE9KChIkpSenm6zPD093VoXFBSk48eP26w/f/68Tpw4YbUpiKenp3x9fW1uAAAAAAAAAABcqswW0UNDQxUUFKSkpCRrWVZWljZv3qzw8HBJUnh4uDIyMpSSkmK1Wbt2rfLy8tShQ4dSjxkAAAAAAAAAULG4O/Pgp0+f1g8//GDdP3jwoHbu3KmAgADVr19fI0aM0NSpU9W4cWOFhobqmWeeUXBwsHr37i1Jatq0qXr06KEHHnhA8+fPV05OjuLi4tS/f38FBwc7qVcAAAAAAAAAgIrCqUX0bdu26eabb7bujxo1SpIUExOjRYsWacyYMTpz5owefPBBZWRkqEuXLlq9erW8vLysbZYsWaK4uDh1795drq6u6tevn+bMmVPqfQEAAAAAAAAAVDxOLaLfdNNNMsbYXe/i4qLJkydr8uTJdtsEBARo6dKlJREeAAAAAAAAAKCSK7NzogMAAAAAAAAA4GwU0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADoroAAAAAAAAAADYQREdAAAAAAAAAAA7KKIDAAAAAAAAAGAHRXQAAAAAAAAAAOygiA4AAAAAAAAAgB0U0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADoroAAAAAAAAAADYQREdAAAAAAAAAAA7KKIDAAAAAAAAAGAHRXQAAAAAAAAAAOygiA4AAAAAAAAAgB0U0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADoroAAAAAAAAAADYQREdAAAAAAAAAAA7KKIDAAAAAAAAAGAHRXQAAAAAAAAAAOygiA4AAAAAAAAAgB0U0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADoroAAAAAAAAAADYQREdAAAAAAAAAAA7KKIDAAAAAAAAAGCHu7MDAAAAAABcmQZjVzk7hHLl0PQoZ4cAAADKEa5EBwAAAAAAAADADoroAAAAAAAAAADYQREdAAAAAAAAAAA7KKIDAAAAAAAAAGAHRXQAAAAAAAAAAOygiA4AAAAAAAAAgB0U0QEAAAAAAAAAsIMiOgAAAAAAAAAAdlBEBwAAAAAAAADADndnBwCg/GswdpWzQyjQoelRzg4BAAAAAAAA5RxXogMAAAAAAAAAYAdFdAAAAAAAAAAA7KCIDgAAAAAAAACAHRWmiD537lw1aNBAXl5e6tChg7Zs2eLskAAAQBGQywEAKN/I5QCAiqpCFNHfffddjRo1ShMmTND27dvVqlUrRUZG6vjx484ODQAAFAK5HACA8o1cDgCoyCpEEX3mzJl64IEHNHjwYIWFhWn+/PmqWrWq3nzzTWeHBgAACoFcDgBA+UYuBwBUZO7ODuBKnTt3TikpKRo3bpy1zNXVVREREUpOTnZiZACAK9Vg7Cpnh1CgQ9OjnB1ChUIuBwCgfCOXAwAqunJfRP/tt9+Um5urwMBAm+WBgYH6/vvvC9wmOztb2dnZ1v3MzExJUlZWlkNiysv+wyH7AXBlHPWahvOU1b+njnpuXdiPMcYh+yuvylouL6vPOwCA45DLHauouZzP5ACAK1XaubzcF9GLIz4+XpMmTcq3vF69ek6IBkBJ8Zvt7AhQUTn6uXXq1Cn5+fk5dqcVHLkcAHAlyOXORR4HAFyp0s7l5b6IXqtWLbm5uSk9Pd1meXp6uoKCggrcZty4cRo1apR1Py8vTydOnFDNmjXl4uJSovGWZVlZWapXr56OHDkiX19fZ4dTZjFOl8cYFQ7jVDgVeZyMMTp16pSCg4OdHYpTlaVcXpGfb2UdY+88jL3zMPbO4chxJ5f/pai5vCQ/k1fm11Vl7rtUuftP3+k7fS++wubycl9E9/DwUNu2bZWUlKTevXtL+isBJyUlKS4ursBtPD095enpabPM39+/hCMtP3x9fSvdi684GKfLY4wKh3EqnIo6Tly1VjZzeUV9vpUHjL3zMPbOw9g7h6PGnVxe9FxeGp/JK/PrqjL3Xarc/afv9L2yKc1cXu6L6JI0atQoxcTEqF27dmrfvr1mz56tM2fOaPDgwc4ODQAAFAK5HACA8o1cDgCoyCpEEf2ee+7Rr7/+qvHjxystLU2tW7fW6tWr8/2oCQAAKJvI5QAAlG/kcgBARVYhiuiSFBcXZ/cr3ygcT09PTZgwId/X6mCLcbo8xqhwGKfCYZwqj7KQy3m+OQ9j7zyMvfMw9s7BuJcccrlzVea+S5W7//Sdvlc2zui7izHGlNrRAAAAAAAAAAAoR1ydHQAAAAAAAAAAAGUVRXQAAAAAAAAAAOygiA4AACqVQYMGqXfv3qVyrAYNGmj27NmlcqzyZtGiRfL393d2GECF8dVXX8nFxUUZGRnODqXc4+8TAAC4FEX0Sm769OlycXHRiBEjrGVnz55VbGysatasqWrVqqlfv35KT093XpBO8ssvv+i+++5TzZo15e3trRYtWmjbtm3WemOMxo8frzp16sjb21sRERE6cOCAEyMufbm5uXrmmWcUGhoqb29vNWzYUFOmTNHFP7VQ2cZpw4YNuuOOOxQcHCwXFxd99NFHNusLMx4nTpxQdHS0fH195e/vr6FDh+r06dOl2IuS93fjlJOToyeffFItWrSQj4+PgoODNXDgQB09etRmH5VhnCqLtLQ0DR8+XI0aNZKXl5cCAwPVuXNnzZs3T3/88YezwysUewWXrVu36sEHHyz9gMqIQYMGycXFRS4uLvLw8FCjRo00efJknT9/vlDbT5w40dre3d1dtWrV0o033qjZs2crOzu7hKMvey6M58MPP5xvXWxsrFxcXDRo0CCHHKugHFYZxMfH6/rrr1f16tVVu3Zt9e7dW/v27XPIvn/44QcNHjxYdevWlaenp0JDQ3XvvffavL9E4Vz8t8XFxUU1a9ZUjx499O233zpk//fcc4/279/vkH2hZMydO1cNGjSQl5eXOnTooC1btvxt++XLl6tJkyby8vJSixYt9Nlnn5VSpI5XlL4vWrTI5rXi4uIiLy+vUozWcS73OasgX331la677jp5enqqUaNGWrRoUYnHWRKK2vcLJ1UvvaWlpZVOwA5U3LxcEV7zxel7RXnNz5s3Ty1btpSvr698fX0VHh6uzz///G+3KY3HnCJ6JbZ161a99tpratmypc3ykSNH6tNPP9Xy5cu1fv16HT16VH379nVSlM5x8uRJde7cWVWqVNHnn3+uPXv26MUXX1SNGjWsNjNmzNCcOXM0f/58bd68WT4+PoqMjNTZs2edGHnpeu655zRv3jy98sor2rt3r5577jnNmDFDL7/8stWmso3TmTNn1KpVK82dO7fA9YUZj+joaO3evVuJiYlauXKlNmzYUOGKcH83Tn/88Ye2b9+uZ555Rtu3b9eHH36offv26Z///KdNu8owTpXBTz/9pDZt2mjNmjWaNm2aduzYoeTkZI0ZM0YrV67Ul19+WeB2OTk5pRxp8Vx11VWqWrWqs8Nwqh49eujYsWM6cOCAHn/8cU2cOFHPP/98obdv1qyZjh07ptTUVK1bt0533XWX4uPj1alTJ506daoEIy+b6tWrp2XLlunPP/+0lp09e1ZLly5V/fr1nRhZwcrLa/WC9evXKzY2Vps2bVJiYqJycnJ066236syZM1e0323btqlt27bav3+/XnvtNe3Zs0crVqxQkyZN9Pjjjzso+srlwt+WY8eOKSkpSe7u7rr99tsdsm9vb2/Vrl3bIfuC47377rsaNWqUJkyYoO3bt6tVq1aKjIzU8ePHC2y/ceNG3XvvvRo6dKh27Nih3r17q3fv3vruu+9KOfIrV9S+S5Kvr6/1Wjl27JgOHz5cihE7zuU+Z13q4MGDioqK0s0336ydO3dqxIgRuv/++/XFF1+UcKSOV9S+X7Bv3z6bx748/l0rTl6uKK/54r4nqQiv+bp162r69OlKSUnRtm3b1K1bN/Xq1Uu7d+8usH2pPeYGldKpU6dM48aNTWJiounatasZPny4McaYjIwMU6VKFbN8+XKr7d69e40kk5yc7KRoS9+TTz5punTpYnd9Xl6eCQoKMs8//7y1LCMjw3h6epr//Oc/pRFimRAVFWWGDBlis6xv374mOjraGMM4STIrVqyw7hdmPPbs2WMkma1bt1ptPv/8c+Pi4mJ++eWXUou9NF06TgXZsmWLkWQOHz5sjKmc41RRRUZGmrp165rTp08XuD4vL88Y89fz5NVXXzV33HGHqVq1qpkwYYI5f/68GTJkiGnQoIHx8vIy11xzjZk9e7bN9ufPnzcjR440fn5+JiAgwIwePdoMHDjQ9OrVy2oTEhJiZs2aZbNdq1atzIQJE6z7L774omnevLmpWrWqqVu3rhk2bJg5deqUMcaYdevWGUk2twvbXrrvw4cPm3/+85/Gx8fHVK9e3dx1110mLS3NWj9hwgTTqlUr89Zbb5mQkBDj6+tr7rnnHpOVlVXEkS0bYmJibMbaGGNuueUW07FjR5OQkGD8/PzM6tWrTZMmTYyPj4+JjIw0R48etdpeGI9L7d2713h4eJinnnqqhHtQtlwYz+bNm5t33nnHWr5kyRLTsmVL06tXLxMTE2OM+etvYufOna3nflRUlPnhhx+sbbKzs01sbKwJCgoynp6epn79+mbatGnGmL+etxc/n0NCQqztPvroI9OmTRvj6elpQkNDzcSJE01OTo61vqDXanl2/PhxI8msX7/eWibJvP7666Z3797G29vbNGrUyHz88cd295GXl2eaNWtm2rZta3Jzc/OtP3nypDHGmJtvvtnExsbmO36VKlXMl19+aYwx5uzZs2bMmDGmbt26xsPDwzRs2NC88cYbxpj/+1t0YX/GGPPf//7XdOnSxXh5eZm6deuaRx991O7f2/KkoL8t//3vf40kc/z4cWOMMWPGjDGNGzc23t7eJjQ01Dz99NPm3LlzVvudO3eam266yVSrVs1Ur17dXHfdddb7igt/ny72ySefmHbt2hlPT09Ts2ZN07t37xLtI+xr3769zWslNzfXBAcHm/j4+ALb33333SYqKspmWYcOHcxDDz1UonGWhKL2vaDnckVQmM8PY8aMMc2aNbNZds8995jIyMgSjKzkFabvBeWDiqKgvHypivSav1hh+l5RX/PGGFOjRg3rPc+lSusx50r0Sio2NlZRUVGKiIiwWZ6SkqKcnByb5U2aNFH9+vWVnJxc2mE6zSeffKJ27drprrvuUu3atdWmTRu9/vrr1vqDBw8qLS3NZpz8/PzUoUOHSjVOnTp1UlJSkvV112+++Ub/+9//1LNnT0mM06UKMx7Jycny9/dXu3btrDYRERFydXXV5s2bSz3msiIzM1MuLi7WdBmMU8Xw+++/a82aNYqNjZWPj0+BbVxcXKz/T5w4UX369NGuXbs0ZMgQ5eXlqW7dulq+fLn27Nmj8ePH69///rfee+89a5sXX3xRixYt0ptvvqn//e9/OnHihFasWFHkWF1dXTVnzhzt3r1bixcv1tq1azVmzBhJf/0tnD17ts1VH0888US+feTl5alXr146ceKE1q9fr8TERP3000+65557bNr9+OOP+uijj7Ry5UqtXLlS69ev1/Tp04scc1nl7e2tc+fOSfrrmycvvPCC3n77bW3YsEGpqakFjt2lmjRpop49e+rDDz8s6XDLpCFDhighIcG6/+abb2rw4ME2bc6cOaNRo0Zp27ZtSkpKkqurq/r06aO8vDxJ0pw5c/TJJ5/ovffe0759+7RkyRI1aNBA0l/fVpSkhIQEHTt2zLr/3//+VwMHDtTw4cO1Z88evfbaa1q0aJGeffZZm2Nf+lotzzIzMyVJAQEBNssnTZqku+++W99++61uu+02RUdH68SJEwXuY+fOndq9e7cef/xxubrm//h1Ibfdf//9Wrp0qc1URe+8847+8Y9/qFu3bpKkgQMH6j//+Y/mzJmjvXv36rXXXlO1atUKPO6PP/6oHj16qF+/fvr222/17rvv6n//+5/i4uKKPA5l3enTp/XOO++oUaNGqlmzpiSpevXqWrRokfbs2aOXXnpJr7/+umbNmmVtEx0drbp162rr1q1KSUnR2LFjVaVKlQL3v2rVKvXp00e33XabduzYoaSkJLVv375U+gZb586dU0pKis37aVdXV0VERNj9fJGcnJzvc29kZGS5+zxSnL5Lf70+QkJCVK9evb+9krOiqSiP+5Vo3bq16tSpo1tuuUVff/21s8NxCHt5+WIV9bEvTN+liveaz83N1bJly3TmzBmFh4cX2Ka0HnN3h+4N5cKyZcu0fft26wPRxdLS0uTh4ZFvXtfAwMByOX9Wcf3000+aN2+eRo0apX//+9/aunWrHnvsMXl4eCgmJsYai8DAQJvtKts4jR07VllZWWrSpInc3NyUm5urZ599VtHR0ZLEOF2iMOORlpaW72t27u7uCggIqJRjJv01TcGTTz6pe++9V76+vpIYp4rihx9+kDFG1157rc3yWrVqWVMcxcbG6rnnnpMk/etf/8pXKJw0aZL1/9DQUCUnJ+u9997T3XffLUmaPXu2xo0bZ01LNn/+/GJ9jffi3w5p0KCBpk6dqocfflivvvqqPDw85OfnJxcXFwUFBdndR1JSknbt2qWDBw+qXr16kqS33npLzZo109atW3X99ddL+qvYvmjRIlWvXl2SNGDAACUlJeUrVJY3xhglJSXpiy++0KOPPirpr6k+5s+fr4YNG0qS4uLiNHny5ELtr0mTJlqzZk2JxVuW3XfffRo3bpz19dyvv/5ay5Yt01dffWW16devn802b775pq666irt2bNHzZs3V2pqqho3bqwuXbrIxcVFISEhVturrrpK0l/F3Yuf05MmTdLYsWMVExMjSbr66qs1ZcoUjRkzRhMmTLDaFfRaLY/y8vI0YsQIde7cWc2bN7dZN2jQIN17772SpGnTpmnOnDnasmWLevTokW8/F377pEmTJn97vL59+youLk4ff/yx9Tds0aJF1vzf+/fv13vvvafExETrg+LVV19td3/x8fGKjo62/n41btxYc+bMUdeuXTVv3rxyOUfqxVauXGmdQDhz5ozq1KmjlStXWicqnn76aattgwYN9MQTT2jZsmXWCdDU1FSNHj3aelwaN25s91jPPvus+vfvb5NzWrVq5fA+4fJ+++035ebmFvh++vvvvy9wm7S0tArxeaQ4fb/22mv15ptvqmXLlsrMzNQLL7ygTp06affu3apbt25phO009h73rKws/fnnn/L29nZSZCWvTp06mj9/vtq1a6fs7Gy98cYbuummm7R582Zdd911zg6v2P4uL1+sorzmL1bYvlek1/yuXbsUHh6us2fPqlq1alqxYoXCwsIKbFtajzlF9ErmyJEjGj58uBITE8v9G+eSlJeXp3bt2mnatGmSpDZt2ui7777T/PnzrQ+OkN577z0tWbJES5cuVbNmzay55oKDgxknOEROTo7uvvtuGWM0b948Z4eDUrJlyxbl5eUpOjra5orMi795cMHcuXP15ptvKjU1VX/++afOnTun1q1bS/rrao1jx46pQ4cOVnt3d3e1a9fO5geQC+PLL79UfHy8vv/+e2VlZen8+fM6e/as/vjjj0LPeb53717Vq1fPKqBLUlhYmPz9/bV3716riN6gQQOrgC799UHo7+Y6LesuFLpycnKUl5enf/3rX5o4caKWL1+uqlWrWgV0qWh9NcbYfFOhMrnqqqsUFRWlRYsWyRijqKgo1apVy6bNgQMHNH78eG3evFm//fabdQV6amqqmjdvrkGDBumWW27Rtddeqx49euj222/Xrbfe+rfH/eabb/T111/bnNDJzc3N91oo6LVaHsXGxuq7777T//73v3zrLv5NIR8fH/n6+tp97hb2742Xl5cGDBigN998U3fffbe2b9+u7777Tp988omkv65od3NzU9euXQu1v2+++UbffvutlixZYhNLXl6eDh48qKZNmxZqP2XVzTffbL03OHnypF599VX17NlTW7ZsUUhIiN59913NmTNHP/74o06fPq3z589bJ+MladSoUbr//vv19ttvKyIiQnfddZfN36OL7dy5Uw888ECp9AtwpPDwcJsrNzt16qSmTZvqtdde05QpU5wYGUrStddea3ORSqdOnfTjjz9q1qxZevvtt50Y2ZX5u7xc0RW27xXpNX/ttddq586dyszM1Pvvv6+YmBitX7/ebiG9NDCdSyWTkpKi48eP67rrrpO7u7vc3d21fv16zZkzR+7u7goMDNS5c+eUkZFhs116evrfXl1X0dSpUyffC7Np06ZKTU2VJGss0tPTbdpUtnEaPXq0xo4dq/79+6tFixYaMGCARo4cqfj4eEmM06UKMx5BQUH5PoSfP39eJ06cqHRjdqGAfvjwYSUmJtp88GWcKoZGjRrJxcUl36/MX3311WrUqFG+K4QunfJl2bJleuKJJzR06FCtWbNGO3fu1ODBg62pQgrL1dU1X5Hr4h9DPHTokG6//Xa1bNlSH3zwgVJSUqwfdirqsQrj0ukEXFxcrAJoeXThB70OHDigP//8U4sXL7Yey4L6WtiC4969exUaGurweMuLIUOGaNGiRVq8eHGBU6bccccdOnHihF5//XVt3rzZmurqwnP2uuuu08GDBzVlyhT9+eefuvvuu3XnnXf+7TFPnz6tSZMmaefOndZt165dOnDggM3FGfamZypP4uLitHLlSq1bt67AK7eK8jq95pprJMnulaIXu//++5WYmKiff/5ZCQkJ6tatm/UtgaJeNXn69Gk99NBDNo/XN998owMHDtgtFpcnPj4+atSokRo1aqTrr79eb7zxhs6cOaPXX39dycnJio6O1m233aaVK1dqx44deuqpp2z+Zk+cOFG7d+9WVFSU1q5dq7CwMLvTfVXkK1bLm1q1asnNza1Iny+CgoIqxOeR4vT9UlWqVFGbNm30ww8/lESIZYq9x93X17dSvqbbt29frh/3y+Xli1WU1/wFRen7pcrza97Dw0ONGjVS27ZtFR8fr1atWumll14qsG1pPeYU0SuZ7t27a9euXTZvptu1a6fo6Gjr/1WqVFFSUpK1zb59+5Sammp37qGKqHPnzvmKOvv377c+xISGhiooKMhmnLKysrR58+ZKNU5//PFHvrk93dzcrA+RjJOtwoxHeHi4MjIylJKSYrVZu3at8vLybK6mreguFNAPHDigL7/80prf9ALGqWKoWbOmbrnlFr3yyiuX/ZX5gnz99dfq1KmTHnnkEbVp00aNGjXSjz/+aK338/NTnTp1bObJP3/+vM3zRvrrqt5jx45Z97OysnTw4EHrfkpKivLy8vTiiy+qY8eOuuaaa3T06FGbfXh4eCg3N/dv423atKmOHDmiI0eOWMv27NmjjIwMp15RUdIuFLrq168vd3fHfAny+++/1+rVq/NNWVKZ9OjRQ+fOnVNOTo4iIyNt1v3+++/at2+fnn76aXXv3l1NmzbVyZMn8+3D19dX99xzj15//XW9++67+uCDD6x5vatUqZLvOX3ddddp3759VuHy4ltBc32XR8YYxcXFacWKFVq7dq1DTtS0bt1aYWFhevHFFwsstF988UqLFi3Url07vf7661q6dKnNCZIWLVooLy9P69evL9Rxr7vuOu3Zs6fAx8vDw+OK+1XWuLi4yNXVVX/++ac2btyokJAQPfXUU2rXrp0aN25sTX90sWuuuUYjR47UmjVr1LdvX5vfGrhYy5Ytbd6/wXk8PDzUtm1bm8cjLy9PSUlJdj9fhIeH53v8EhMTy93nkeL0/VK5ubnatWuX6tSpU1JhlhkV5XF3lJ07d5bLx704ebmiPPaOeE9SkV7zeXl5Nt9SvlhpPeZM51LJVK9ePd/8ST4+PqpZs6a1fOjQoRo1apQCAgLk6+urRx99VOHh4erYsaMzQnaKkSNHqlOnTpo2bZruvvtubdmyRQsWLNCCBQsk/fUmfcSIEZo6daoaN26s0NBQPfPMMwoODlbv3r2dG3wpuuOOO/Tss8+qfv36atasmXbs2KGZM2daH/gq4zidPn3a5izvwYMHtXPnTgUEBKh+/fqXHY+mTZuqR48eeuCBBzR//nzl5OQoLi5O/fv3V3BwsJN65Xh/N0516tTRnXfeqe3bt2vlypXKzc215jILCAiQh4dHpRmnyuDVV19V586d1a5dO02cOFEtW7aUq6urtm7dqu+//15t27a1u23jxo311ltv6YsvvlBoaKjefvttbd261eYN5vDhwzV9+nQ1btxYTZo00cyZM/N926pbt25atGiR7rjjDvn7+2v8+PFyc3Oz1jdq1Eg5OTl6+eWXdccdd+jrr7/W/PnzbfbRoEEDnT59WklJSWrVqpWqVq2ab5qXiIgItWjRQtHR0Zo9e7bOnz+vRx55RF27dq0w01+UhPPnzystLU15eXn6/fff9dVXX2nq1Klq3bq1Ro8e7ezwnMbNzU179+61/n+xGjVqqGbNmlqwYIHq1Kmj1NRUjR071qbNzJkzVadOHbVp00aurq5avny5goKCrN/FadCggZKSktS5c2d5enqqRo0aGj9+vG6//XbVr19fd955p1xdXfXNN9/ou+++09SpU0ul3yUtNjZWS5cu1ccff6zq1atb+cfPz6/YVy66uLgoISFBERERuuGGG/TUU0+pSZMmOn36tD799FOtWbPGpjB+//33Ky4uTj4+PurTp4+1vEGDBoqJidGQIUM0Z84ctWrVSocPH9bx48etOdQv9uSTT6pjx46Ki4vT/fffLx8fH+3Zs0eJiYl65ZVXitWXsiQ7O9t6fE6ePKlXXnlFp0+f1h133KGsrCylpqZq2bJluv7667Vq1Sqbq8z//PNPjR49WnfeeadCQ0P1888/a+vWrXZPzE2YMEHdu3dXw4YN1b9/f50/f16fffaZnnzyyVLpK2yNGjVKMTExateundq3b6/Zs2frzJkz1m8xDBw4UP/4xz+sb8cOHz5cXbt21YsvvqioqCgtW7ZM27Ztsz7blSdF7fvkyZPVsWNHNWrUSBkZGXr++ed1+PBh3X///c7sRrFc7nPWuHHj9Msvv+itt96SJD388MN65ZVXNGbMGA0ZMkRr167Ve++9p1WrVjmrC8VW1L7Pnj1boaGhatasmc6ePas33nhDa9euLZe/JVOYvFxRX/PF6XtFec2PGzdOPXv2VP369XXq1CktXbpUX331lfXbVk57zA0qva5du5rhw4db9//880/zyCOPmBo1apiqVauaPn36mGPHjjkvQCf59NNPTfPmzY2np6dp0qSJWbBggc36vLw888wzz5jAwEDj6elpunfvbvbt2+ekaJ0jKyvLDB8+3NSvX994eXmZq6++2jz11FMmOzvbalPZxmndunVGUr5bTEyMMaZw4/H777+be++911SrVs34+vqawYMHm1OnTjmhNyXn78bp4MGDBa6TZNatW2ftozKMU2Vx9OhRExcXZ0JDQ02VKlVMtWrVTPv27c3zzz9vzpw5Y4wxRpJZsWKFzXZnz541gwYNMn5+fsbf398MGzbMjB071rRq1cpqk5OTY4YPH258fX2Nv7+/GTVqlBk4cKDp1auX1SYzM9Pcc889xtfX19SrV88sWrTItGrVykyYMMFqM3PmTFOnTh3j7e1tIiMjzVtvvWUkmZMnT1ptHn74YVOzZk0jydo2JCTEzJo1y2pz+PBh889//tP4+PiY6tWrm7vuusukpaVZ6ydMmGATvzHGzJo1y4SEhBRjZJ0vJibGZqwvlpCQYPz8/GyWrVixwlz89nTChAnW69/Nzc0EBASYLl26mFmzZpmzZ8+WYORl09+NpzHG9OrVy8o3iYmJpmnTpsbT09O0bNnSfPXVVzavowULFpjWrVsbHx8f4+vra7p37262b99u7euTTz4xjRo1Mu7u7jbPv9WrV5tOnToZb29v4+vra9q3b2/zHqmg12p5Yi//JCQk2LS5tI9+fn42bQqyb98+M3DgQBMcHGw8PDxMSEiIuffee23G3RhjTp06ZapWrWoeeeSRfPv4888/zciRI02dOnWMh4eHadSokXnzzTeNMf+XWy/+u7RlyxZzyy23mGrVqhkfHx/TsmVL8+yzzxZpTMqimJgYm8enevXq5vrrrzfvv/++1Wb06NGmZs2aplq1auaee+4xs2bNsv7mZGdnm/79+5t69eoZDw8PExwcbOLi4syff/5pjCn479MHH3xgWrdubTw8PEytWrVM3759S6u7KMDLL79s6tevbzw8PEz79u3Npk2brHVdu3a1/hZe8N5775lrrrnGeHh4mGbNmplVq1aVcsSOU5S+jxgxwmobGBhobrvttnx/c8qLy33OiomJMV27ds23zYXX7dVXX33Zv9NlVVH7/txzz5mGDRsar//X3r3H5nj/fxx/3XX3bC2KaYcWrUNHQ0fjrFJLzbnYFtmoOJOuzKlbiLGty76ow7I5bEoRMeYwZlHFutGGqUPLwqqk4tTNhqEIpZ/fH77un3vtTXXmVt/nI7mSXp/r87nu93U3zfu63/enn8vDw1SrVs1ERkaaH374wTnB/0NlycvP6998ea79efmbHzJkiAkMDDRubm6mRo0aJioqyqSlpdmOO+t3bjHmMZ+sBQAAAAD4V5w6dUoNGjRQVlaWwsPDnR0OAAAAJFFEBwAAAAAnKyoq0sWLFzVx4kTl5+crMzPT2SEBAADgv56PJwABAAAAQAWWmZkpf39/ZWVllXjuAgAAAJyLmegAAAAAAAAAADjATHQAAAAAAAAAABygiA4AAAAAAAAAgAMU0QEAAAAAAAAAcIAiOgAAAAAAAAAADlBEBwAAAAAAAADAAYroAAAAQAVjsVj07bffOjsMDR48WH369HF2GAAAVEhPOp+npKSoSpUqD+3z99wdGRmpcePGPXRMUFCQ5s2b94/jAyoyiugAHNqzZ48qVaqk7t27OzsUAACeW4MHD5bFYimxde3a1dmh2Zw6dUoWi0XZ2dl27fPnz1dKSopTYgIA4EkpLQ8/uE2fPt3hWEc58t+Qnp6ubt26yc/PT15eXgoNDdWECRN07ty5Mp+D3A2UD0V0AA4lJyfrnXfe0a5du3T+/HlnhwMAwHOra9euKigosNtWr17t7LAeydfX95Ez3gAAeNY9mH/nzZsnHx8fu7aJEyc6O0QtXrxYXbp0Ua1atbR+/XodPXpUixYt0pUrV5SUlFTm8/wbubuoqOiJng94FlFEB1CqwsJCrVmzRqNHj1b37t1LfFO9efNmhYSEyMPDQ507d9by5ctlsVj0119/2fpkZGSoQ4cO8vT0VJ06dRQfH6/r168/3QsBAKACcHd3V61atey2qlWrSpLy8vLUsWNHeXh4KDQ0VNu3b7cb++OPP5bIwdnZ2bJYLDp16pStLTMzU5GRkfLy8lLVqlUVHR2ty5cvS5JSU1PVvn17ValSRX5+furRo4dOnjxpG1uvXj1JUosWLWSxWBQZGSmp5L+E37p1S/Hx8apZs6Y8PDzUvn17ZWVllYh1586datmypby8vNS2bVvl5uY+ibcRAIByeTD/+vr6ymKx2PZr1qypOXPmqHbt2nJ3d1fz5s2VmppqG+soR2ZlZenVV19V9erV5evrq06dOungwYPliu/s2bOKj49XfHy8li5dqsjISAUFBaljx45asmSJpk2bZtd/27ZtatKkiSpXrmz7ov6+Ry3FduHCBfXs2VOenp6qV6+eVq1aVaKPxWLRwoUL1atXL3l7eysxMVGStGnTJoWHh8vDw0P169fXjBkzdOfOHbtxS5YsUUxMjLy8vBQSEqLNmzeX6z0BnjaK6ABKtXbtWjVu3FiNGjXS22+/raVLl8oYI0nKz89X//791adPH+Xk5GjkyJGaMmWK3fiTJ0+qa9eu6tevnw4fPqw1a9YoIyNDcXFxzrgcAAAqpOLiYvXt21dubm76+eeftWjRIiUkJDz2ebKzsxUVFaXQ0FDt2bNHGRkZ6tmzp+7evStJun79usaPH6/9+/dr586dcnFxUUxMjIqLiyVJ+/btkyTt2LFDBQUF2rBhQ6mvM3nyZK1fv17Lly/XwYMHFRwcrOjoaF26dMmu35QpU5SUlKT9+/fLarVqyJAhj31NAAA8DfPnz1dSUpJmz56tw4cPKzo6Wr169VJeXp4kxzny2rVrio2NVUZGhvbu3auQkBB169ZN165de+wYvvnmG92+fVuTJ08u9fiDM8tv3Lih2bNna+XKldq1a5dOnz79WDPpBw8erDNnzig9PV3r1q3TggULdOHChRL9pk+frpiYGB05ckRDhgzR7t27NWjQII0dO1ZHjx7V4sWLlZKSYiuw3zdjxgy98cYbOnz4sLp166a33nqrxH0C8EwyAFCKtm3bmnnz5hljjCkqKjLVq1c36enpxhhjEhISTNOmTe36T5kyxUgyly9fNsYYM3ToUDNixAi7Prt37zYuLi7m5s2b/3r8AABUFLGxsaZSpUrG29vbbktMTDTbtm0zVqvVnDt3ztZ/69atRpLZuHGjMcaY9PR0uxxsjDGHDh0ykkx+fr4xxpgBAwaYdu3alTmmP/74w0gyR44cMcYYk5+fbySZQ4cOlYi9d+/exhhjCgsLjaurq1m1apXt+O3bt01AQICZOXOmXaw7duyw9fn++++NJO4PAADPhGXLlhlfX1/bfkBAgElMTLTr06pVKzNmzBhjjOMc+Xd37941L7zwgvnuu+9sbQ/m84cZPXq08fHxKVPsksyJEydsbV988YV58cUXbfsP5m5jjOnUqZMZO3asMcaY3NxcI8ns27fPdvzYsWNGkpk7d65d3OPGjbN77aioKPPJJ5/Yta1cudL4+/vbjZs6daptv7Cw0EgyW7dufeS1Ac7GTHQAJeTm5mrfvn0aMGCAJMlqterNN99UcnKy7XirVq3sxkRERNjt5+TkKCUlRZUrV7Zt0dHRKi4uVn5+/tO5EAAAKojOnTsrOzvbbhs1apSOHTumOnXqKCAgwNa3TZs2j33++zPRHcnLy9OAAQNUv359+fj4KCgoSJJ0+vTpMr/GyZMnVVRUpHbt2tnaXF1dFRERoWPHjtn1DQsLs/3s7+8vSaXOcgMAwJmuXr2q8+fP2+U2SWrXrl2J3PZ3v//+u4YPH66QkBD5+vrKx8dHhYWFj5Vb7zPGyGKxlKmvl5eXGjRoYNv39/cvc449duyYrFarXnnlFVtb48aNS11DvWXLlnb7OTk5+vDDD+1qAMOHD1dBQYFu3Lhh6/fgPYC3t7d8fHy4B0CFYHV2AACePcnJybpz547dB3ZjjNzd3fX555+X6RyFhYUaOXKk4uPjSxyrW7fuE4sVAIDngbe3t4KDg8s11sXl3rwY899l16SSD/jy9PR86Dl69uypwMBAffXVVwoICFBxcbGaNm2q27dvlyumR3F1dbX9fL8ocH/pGAAAngexsbG6ePGi5s+fr8DAQLm7u6tNmzblyq0NGzbUlStXVFBQYPvy2ZEHc6x0L88+eI/wpHh7e9vtFxYWasaMGerbt2+Jvh4eHg+Nj3sAVATMRAdg586dO1qxYoWSkpLsZsPl5OQoICBAq1evVqNGjbR//367cQ8+NEySwsPDdfToUQUHB5fY3NzcnuYlAQBQYTVp0kRnzpyxeyDY3r177frUqFFDkuz6ZGdn2/UJCwvTzp07S32NixcvKjc3V1OnTlVUVJSaNGlie+Doffdz9/011EvToEEDubm5KTMz09ZWVFSkrKwshYaGPuQqAQB4Nvn4+CggIMAut0n3HtZ9P7c5ypGZmZmKj49Xt27d9PLLL8vd3V1//vlnueLo37+/3NzcNHPmzFKPP/hw8X+icePGunPnjg4cOGBry83NLdP5w8PDlZubW2oN4P4X/kBFxkx0AHa2bNmiy5cva+jQofL19bU71q9fPyUnJ2vt2rWaM2eOEhISNHToUGVnZyslJUXS/88mS0hIUOvWrRUXF6dhw4bJ29tbR48e1fbt28s8mx0AgP8Vt27d0m+//WbXZrVa1aVLFzVs2FCxsbGaNWuWrl69WuJh3sHBwapTp46mT5+uxMREHT9+XElJSXZ93n//fTVr1kxjxozRqFGj5ObmpvT0dL3++uuqVq2a/Pz89OWXX8rf31+nT5/We++9Zze+Zs2a8vT0VGpqqmrXri0PD48S9wne3t4aPXq0Jk2apGrVqqlu3bqaOXOmbty4oaFDhz7BdwsAgKdn0qRJ+uCDD9SgQQM1b95cy5YtU3Z2tlatWiXJcY4MCQnRypUr1bJlS129elWTJk165H+GOVKnTh3NnTtXcXFxunr1qgYNGqSgoCCdPXtWK1asUOXKlUvk/vJo1KiRunbtqpEjR2rhwoWyWq0aN25cmeKeNm2aevToobp166p///5ycXFRTk6OfvnlF3388cf/ODbA2fgqCICd5ORkdenSpcQHY+leEX3//v26du2a1q1bpw0bNigsLEwLFy60faB3d3eXdG/G208//aTjx4+rQ4cOatGihaZNm2a3RAwAALgnNTVV/v7+dlv79u3l4uKijRs36ubNm4qIiNCwYcOUmJhoN9bV1VWrV6/Wr7/+qrCwMP3nP/8p8WG1YcOGSktLU05OjiIiItSmTRtt2rRJVqtVLi4u+vrrr3XgwAE1bdpU7777rmbNmmU33mq16rPPPtPixYsVEBCg3r17l3odn376qfr166eBAwcqPDxcJ06c0LZt21S1atUn+4YBAPCUxMfHa/z48ZowYYKaNWum1NRUbd68WSEhIZIc58jk5GRdvnxZ4eHhGjhwoOLj41WzZs1yxzFmzBilpaXp3LlziomJUePGjTVs2DD5+Pho4sSJT+RaJWnZsmUKCAhQp06d1LdvX40YMaJMcUdHR2vLli1KS0tTq1at1Lp1a82dO1eBgYFPLDbAmSzm31gYCcD/nMTERC1atEhnzpxxdigAAAAAAADAE8NyLgDKZcGCBWrVqpX8/PyUmZmpWbNmKS4uztlhAQAAAAAAAE8Uy7kAKJe8vDz17t1boaGh+uijjzRhwgRNnz7d2WEBAAAAAFChfPLJJ6pcuXKp22uvvebs8ACI5VwAAAAAAAAAp7l06ZIuXbpU6jFPT0+99NJLTzkiAH9HER0AAAAAAAAAAAdYzgUAAAAAAAAAAAcoogMAAAAAAAAA4ABFdAAAAAAAAAAAHKCIDgAAAAAAAACAAxTRAQAAAAAAAABwgCI6AAAAAAAAAAAOUEQHAAAAAAAAAMABiugAAAAAAAAAADjwf8M+ailn0SndAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1500x500 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from datetime import datetime\n",
    "\n",
    "current_year = datetime.now().year\n",
    "# calculate current age of customers and add it to the dataframe\n",
    "df['Age'] = current_year - df['Year_Birth']\n",
    "\n",
    "#visualise the demographics from the table: Age, Education, Kids in the household\n",
    "\n",
    "# create a new column \"Total_Children\" by adding together \"Kidhome\" and \"Teenhome\"\n",
    "df['Total_Children'] = df['Kidhome'] + df['Teenhome']\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# select columns\n",
    "columns_to_plot = ['Age', 'Education', 'Total_Children']\n",
    "bins = [10, 5, 3]\n",
    "\n",
    "# create a figure with subplots\n",
    "fig, axes = plt.subplots(1, len(columns_to_plot), figsize=(15, 5))\n",
    "\n",
    "# iterate through columns and create histograms\n",
    "for i, (column, num_bins) in enumerate(zip(columns_to_plot, bins)):\n",
    "    axes[i].hist(df[column], bins=num_bins)\n",
    "\n",
    "    # add labels and title for each subplot\n",
    "    axes[i].set_xlabel(column)\n",
    "    axes[i].set_ylabel('Frequency')\n",
    "    axes[i].set_title(f'Histogram of {column}')\n",
    "\n",
    "# adjust spacing between subplots\n",
    "plt.tight_layout()\n",
    "\n",
    "# show the figure\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "NczQACZp3RfD"
   },
   "source": [
    "<h4>Time period covered</h4>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "amCVo-_p3TbT",
    "outputId": "41c7d7c3-de32-42c3-bbe9-4c6a3debe19b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Minimum Registration Date: 2020-01-01 00:00:00\n",
      "Maximum Registration Date: 2023-12-08 00:00:00\n",
      "Registration Date Range: 1437 days 00:00:00\n"
     ]
    }
   ],
   "source": [
    "# convert the 'Dt_Customer' column to a datetime object\n",
    "df['Dt_Customer'] = pd.to_datetime(df['Dt_Customer'])\n",
    "\n",
    "# calculate summary statistics\n",
    "min_registration_date = df['Dt_Customer'].min()\n",
    "max_registration_date = df['Dt_Customer'].max()\n",
    "registration_date_range = max_registration_date - min_registration_date\n",
    "\n",
    "print(\"Minimum Registration Date:\", min_registration_date)\n",
    "print(\"Maximum Registration Date:\", max_registration_date)\n",
    "print(\"Registration Date Range:\", registration_date_range)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "bDWkrl822Xam"
   },
   "source": [
    "<h4>1.3.2 Check for missing values</h4>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "RK1zesg92YMP",
    "outputId": "49169e9f-797c-4d08-e4ab-e13a1cc4d21b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Missing Value Counts:\n",
      "ID                      0\n",
      "Year_Birth              0\n",
      "Education               0\n",
      "Marital_Status          0\n",
      "Income                 24\n",
      "Kidhome                 0\n",
      "Teenhome                0\n",
      "Dt_Customer             0\n",
      "Recency                 0\n",
      "MntWines                0\n",
      "MntFruits               0\n",
      "MntMeatProducts         0\n",
      "MntFishProducts         0\n",
      "MntSweetProducts        0\n",
      "MntGoldProds            0\n",
      "NumDealsPurchases       0\n",
      "NumWebPurchases         0\n",
      "NumCatalogPurchases     0\n",
      "NumStorePurchases       0\n",
      "NumWebVisitsMonth       0\n",
      "AcceptedCmp3            0\n",
      "AcceptedCmp4            0\n",
      "AcceptedCmp5            0\n",
      "AcceptedCmp1            0\n",
      "AcceptedCmp2            0\n",
      "Complain                0\n",
      "Z_CostContact           0\n",
      "Z_Revenue               0\n",
      "Response                0\n",
      "Age                     0\n",
      "Total_Children          0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "missing_values = df.isnull()\n",
    "\n",
    "# sum the missing values for each column\n",
    "missing_counts = missing_values.sum()\n",
    "\n",
    "# print missing value counts\n",
    "print(\"Missing Value Counts:\")\n",
    "print(missing_counts)\n",
    "\n",
    "# filter the dataframe to select rows where \"Income\" is missing\n",
    "#rows_with_missing_income = df[df['Income'].isna()]\n",
    "\n",
    "# print the rows with missing values\n",
    "#print(rows_with_missing_income)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "IGGYVXXV2dbD"
   },
   "source": [
    "There are 24 missing values in the 'Income' column.\n",
    "We can use machine learning to estimate these values from the other 2216 rows that are complete.\n",
    "Age and education level are useful features to predict income.\n",
    "(I will not use income for the CLV predictions in this excercise, but for a different approach, it could be useful to complete the dataset.\n",
    "I could also delete these rows.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "1Xsk8J1J2_Bz",
    "outputId": "f0379ceb-8b1f-4ba0-c222-4bdfc1816e80"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Rows with estimated income values completed: \n",
      "         ID  Year_Birth   Education Marital_Status  Income  Kidhome  Teenhome  \\\n",
      "2216   1612        1981         PhD         Single   52013        1         0   \n",
      "2217   5250        1943      Master          Widow   59249        0         0   \n",
      "2218   9235        1957  Graduation         Single   55784        1         1   \n",
      "2219   5798        1973      Master       Together   51610        0         0   \n",
      "2220   2863        1970  Graduation         Single   52474        1         2   \n",
      "2221  10339        1954      Master       Together   56448        0         1   \n",
      "2222   5255        1986  Graduation         Single   48400        1         0   \n",
      "2223   8996        1957         PhD        Married   58124        2         1   \n",
      "2224   3117        1955  Graduation         Single   56294        0         1   \n",
      "2225   7187        1969      Master       Together   52629        1         1   \n",
      "2226   8557        1982  Graduation         Single   49419        1         0   \n",
      "2227   1994        1983  Graduation        Married   49164        1         0   \n",
      "2228  10629        1973    2n Cycle        Married   47357        1         0   \n",
      "2229   8720        1978    2n Cycle       Together   46084        0         0   \n",
      "2230   4345        1964    2n Cycle         Single   49649        1         1   \n",
      "2231   1295        1963  Graduation        Married   54257        0         1   \n",
      "2232   8268        1961         PhD        Married   57106        0         1   \n",
      "2233   7281        1959         PhD         Single   57615        0         0   \n",
      "2234   2902        1958  Graduation       Together   55530        1         1   \n",
      "2235   2437        1989  Graduation        Married   47636        0         0   \n",
      "2236   5079        1971  Graduation        Married   52219        1         1   \n",
      "2237   3769        1972         PhD       Together   54305        1         0   \n",
      "2238  10475        1970      Master       Together   52374        0         1   \n",
      "2239   7244        1951  Graduation         Single   57312        2         1   \n",
      "\n",
      "     Dt_Customer  Recency  MntWines  ...  AcceptedCmp4  AcceptedCmp5  \\\n",
      "2216    29.03.23      637        23  ...             0             0   \n",
      "2217    08.10.21      426       532  ...             0             1   \n",
      "2218    08.03.21      721         7  ...             0             0   \n",
      "2219    11.03.20     1035       445  ...             0             0   \n",
      "2220    14.10.20     1187       738  ...             1             0   \n",
      "2221    16.06.20       30       161  ...             0             0   \n",
      "2222    03.04.21      998         5  ...             0             0   \n",
      "2223    20.12.21      416       230  ...             0             0   \n",
      "2224    12.05.23      596       264  ...             0             0   \n",
      "2225    17.11.20     1261       375  ...             0             0   \n",
      "2226    08.07.20     1009        11  ...             0             0   \n",
      "2227    18.11.22      534         5  ...             0             0   \n",
      "2228    17.12.22     1149        25  ...             0             0   \n",
      "2229    22.06.23     1067        32  ...             1             0   \n",
      "2230    26.11.22     1038         5  ...             0             0   \n",
      "2231    29.12.21      709       231  ...             0             0   \n",
      "2232    03.03.22      792       352  ...             0             0   \n",
      "2233    20.09.20     1181        81  ...             0             0   \n",
      "2234    03.04.21     1088        19  ...             0             0   \n",
      "2235    10.10.21     1321       861  ...             1             0   \n",
      "2236    04.06.23      345        71  ...             0             0   \n",
      "2237    21.08.22       67        25  ...             0             0   \n",
      "2238    01.04.22      436       187  ...             0             0   \n",
      "2239    29.07.23      232        48  ...             0             0   \n",
      "\n",
      "      AcceptedCmp1  AcceptedCmp2  Complain  Z_CostContact  Z_Revenue  \\\n",
      "2216             0             0         0              3         11   \n",
      "2217             0             0         0              3         11   \n",
      "2218             0             0         0              3         11   \n",
      "2219             0             0         0              3         11   \n",
      "2220             1             0         0              3         11   \n",
      "2221             0             0         0              3         11   \n",
      "2222             0             0         0              3         11   \n",
      "2223             0             0         0              3         11   \n",
      "2224             0             0         0              3         11   \n",
      "2225             0             0         0              3         11   \n",
      "2226             0             0         0              3         11   \n",
      "2227             0             0         0              3         11   \n",
      "2228             0             0         0              3         11   \n",
      "2229             0             0         0              3         11   \n",
      "2230             0             0         0              3         11   \n",
      "2231             0             0         0              3         11   \n",
      "2232             0             0         0              3         11   \n",
      "2233             0             0         0              3         11   \n",
      "2234             0             0         0              3         11   \n",
      "2235             1             0         0              3         11   \n",
      "2236             0             0         0              3         11   \n",
      "2237             0             0         0              3         11   \n",
      "2238             0             0         0              3         11   \n",
      "2239             0             0         0              3         11   \n",
      "\n",
      "      Response  Age  Total_Children  \n",
      "2216         0   42               1  \n",
      "2217         1   80               0  \n",
      "2218         0   66               2  \n",
      "2219         0   50               0  \n",
      "2220         0   53               3  \n",
      "2221         0   69               1  \n",
      "2222         0   37               1  \n",
      "2223         0   66               3  \n",
      "2224         0   68               1  \n",
      "2225         0   54               2  \n",
      "2226         0   41               1  \n",
      "2227         0   40               1  \n",
      "2228         0   50               1  \n",
      "2229         0   45               0  \n",
      "2230         0   59               2  \n",
      "2231         0   60               1  \n",
      "2232         0   62               1  \n",
      "2233         0   64               0  \n",
      "2234         0   65               2  \n",
      "2235         0   34               0  \n",
      "2236         0   52               2  \n",
      "2237         0   51               1  \n",
      "2238         0   53               1  \n",
      "2239         0   72               3  \n",
      "\n",
      "[24 rows x 31 columns]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.pipeline import Pipeline\n",
    "\n",
    "# separate dataset into complete data and data with missing income\n",
    "complete_data = df[df['Income'].notnull()]\n",
    "eligible_missing_data = df[df['Income'].isnull() & df['Age'].notnull() & df['Education'].notnull()]\n",
    "\n",
    "\n",
    "# check if there is any missing data to predict\n",
    "if not eligible_missing_data.empty:\n",
    "    # select features and target variable\n",
    "    X = complete_data[['Age', 'Education']]\n",
    "    y = complete_data['Income']\n",
    "\n",
    "    # split data into training and validation sets\n",
    "    X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.2, random_state=42)\n",
    "\n",
    "    # define a preprocessor for encoding categorical variables (Education)\n",
    "    preprocessor = ColumnTransformer(\n",
    "        transformers=[\n",
    "            ('edu', OneHotEncoder(), ['Education'])\n",
    "        ],\n",
    "        remainder='passthrough'\n",
    "    )\n",
    "\n",
    "    # create a pipeline with preprocessing and linear regression\n",
    "    pipeline = Pipeline([\n",
    "        ('preprocessor', preprocessor),\n",
    "        ('regressor', LinearRegression())\n",
    "    ])\n",
    "\n",
    "    # train the model\n",
    "    pipeline.fit(X_train, y_train)\n",
    "\n",
    "    # make predictions for the missing income values\n",
    "    eligible_missing_data_copy = eligible_missing_data.copy()\n",
    "    eligible_missing_data_copy['Income'] = pipeline.predict(eligible_missing_data_copy[['Age', 'Education']]).astype(int)\n",
    "\n",
    "    print('Rows with estimated income values completed: ')\n",
    "    print(eligible_missing_data_copy)\n",
    "\n",
    "    # replace the original dataframe with the updated one\n",
    "    df.update(eligible_missing_data_copy)\n",
    "\n",
    "# if missing_data is empty, no predictions are made, and the original dataframe remains unchanged."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "C2ED4zIS3eNh"
   },
   "source": [
    "<h4>1.3.3 Print the number of instances</h4>\n",
    "This step is not needed because it is a regression task, not classification."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "RNh_BG-T3482"
   },
   "source": [
    "<h2>2. Choose a measure of success</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "41jn2cEp354O"
   },
   "source": [
    "**Mean Absolute Error (MAE)**\n",
    "\n",
    "The MAE is the difference between the actual and predicted value, divided by the total number of points in the dataset. A lower value indicates better model performance. The closer the MAE is to 0, the better the model's predictions align with the true values.\n",
    "\n",
    "**Modell loss**\n",
    "\n",
    "A measure that quantifies how well a machine learning model's predictions align with the actual values in the training data or validation data. The loss is a single scalar value that represents the error or discrepancy between the predicted values and the true values for a set of data points."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "GoFmTGWl39R7"
   },
   "source": [
    "<h2>3. Decide on an evaluation protocol</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Ix2mRydw4ANy"
   },
   "source": [
    "\n",
    "I choose  K-fold cross-validation because:\n",
    "- historical CLV is available for a limited period, between 2020-01-01 and 2023-12-08. K-fold cross-validation helps make the most of the available data by partitioning it into training and validation sets multiple times, allowing assessment of model performance more robustly.\n",
    "- Cross-validation helps evaluate how well the model generalizes to unseen data by simulating different data splits, which can help identify overfitting or underfitting issues."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "dL-SjtKD4CQe"
   },
   "source": [
    "<h2>4. Preparing the data</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "KZb2arcU4Egd"
   },
   "source": [
    "Calculate historical CLV\n",
    "\n",
    "There are different opinions and formulas out there how to calculate the customer lifetime value, I used this one from Gartner.com:\n",
    "https://www.gartner.com/en/digital-markets/insights/what-is-customer-lifetime-value\n",
    "\n",
    "\n",
    "Number of purchases / annually + Purchase amount + Length of commercial relationship\n",
    "= CLV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "b5-gqLjS4MRW",
    "outputId": "3940e090-03a1-4474-f80e-67630dc1ebf5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "       ID  Year_Birth   Education Marital_Status    Income  Kidhome  Teenhome  \\\n",
      "0  9432.0        46.0  Graduation       Together  666666.0      1.0       0.0   \n",
      "1  1503.0        47.0         PhD       Together  162397.0      1.0       1.0   \n",
      "2  1501.0        41.0         PhD        Married  160803.0      0.0       0.0   \n",
      "3  5336.0        52.0      Master       Together  157733.0      1.0       0.0   \n",
      "4  8475.0        50.0         PhD        Married  157243.0      0.0       1.0   \n",
      "\n",
      "  Dt_Customer  Recency  MntWines  ...  Z_Revenue  Response   Age  \\\n",
      "0  2020-08-28    830.0       9.0  ...       11.0       0.0  46.0   \n",
      "1  2020-04-21    474.0      85.0  ...       11.0       0.0  47.0   \n",
      "2  2020-02-23    169.0      55.0  ...       11.0       0.0  41.0   \n",
      "3  2022-08-11    196.0      39.0  ...       11.0       0.0  52.0   \n",
      "4  2020-08-31    860.0      20.0  ...       11.0       0.0  50.0   \n",
      "\n",
      "   Total_Children  Total Purchases  Annual Purchases  Total Spending  \\\n",
      "0             1.0             11.0          2.794015            62.0   \n",
      "1             2.0              1.0          0.254001           107.0   \n",
      "2             0.0             44.0         11.176061          1717.0   \n",
      "3             1.0              2.0          0.508003            59.0   \n",
      "4             1.0             37.0          9.398051          1608.0   \n",
      "\n",
      "   Annual Purchase Amount  Length of Commercial Relationship  Historical_CLV  \n",
      "0               15.748086                                  3       55.626305  \n",
      "1               27.178149                                  3       82.296451  \n",
      "2              436.120390                                  3     1341.889353  \n",
      "3               14.986082                                  1       15.494085  \n",
      "4              408.434238                                  3     1253.496868  \n",
      "\n",
      "[5 rows x 37 columns]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-8-0636a703e3f6>:11: FutureWarning: The parsing of 'now' in pd.to_datetime without `utc=True` is deprecated. In a future version, this will match Timestamp('now') and Timestamp.now()\n",
      "  current_year = pd.to_datetime('now').year\n"
     ]
    }
   ],
   "source": [
    "# calculate the annual purchases\n",
    "df['Total Purchases'] = df['NumDealsPurchases'] + df['NumWebPurchases'] + df['NumCatalogPurchases'] + df['NumStorePurchases']\n",
    "total_years = (pd.to_datetime(df['Dt_Customer']).max() - pd.to_datetime(df['Dt_Customer']).min()).days / 365\n",
    "df['Annual Purchases'] = df['Total Purchases'] / total_years\n",
    "\n",
    "# calculate the annual purchase amount\n",
    "df['Total Spending'] = df['MntWines'] + df['MntFruits'] + df['MntMeatProducts'] + df['MntFishProducts'] + df['MntSweetProducts'] + df['MntGoldProds']\n",
    "df['Annual Purchase Amount'] = df['Total Spending'] / total_years\n",
    "\n",
    "# calculate the length of commercial relationship\n",
    "current_year = pd.to_datetime('now').year\n",
    "df['Year_Birth'] = current_year - df['Year_Birth']\n",
    "df['Length of Commercial Relationship'] = current_year - pd.to_datetime(df['Dt_Customer']).dt.year\n",
    "\n",
    "# calculate the historical CLV for each customer\n",
    "df['Historical_CLV'] = (df['Annual Purchases'] + df['Annual Purchase Amount']) * df['Length of Commercial Relationship']\n",
    "\n",
    "print(df.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "socAkcLv4Wdr"
   },
   "source": [
    "- Label encoding\n",
    "- feature selection\n",
    "- feature normalization\n",
    "- split data into training and testing sets\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "D8STi8jf4Th6",
    "outputId": "55f59cda-2321-47f4-c9ef-50c12c9e8222"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0      55.626305\n",
      "1      82.296451\n",
      "2    1341.889353\n",
      "3      15.494085\n",
      "4    1253.496868\n",
      "Name: Historical_CLV, dtype: float64\n",
      "Plot of joint distributions of each of our features along with kernel density estimation plots\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1000x1000 with 20 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.preprocessing import StandardScaler, LabelEncoder\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# label encoding: encode categorical variables (e.g., education)\n",
    "encoder = LabelEncoder()\n",
    "df['Education'] = encoder.fit_transform(df['Education'])\n",
    "\n",
    "# feature selection: split data into features (X) and CLV target (y)\n",
    "#independent variables:\n",
    "X = df[['Annual Purchases', 'Annual Purchase Amount', 'Length of Commercial Relationship', ]]\n",
    "#dependent variable:\n",
    "y = df['Historical_CLV']\n",
    "\n",
    "# standardize/normalize features\n",
    "scaler = StandardScaler()\n",
    "X = scaler.fit_transform(X)\n",
    "\n",
    "# split data into training and testing sets\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)\n",
    "\n",
    "\n",
    "print(y.head())\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "\n",
    "X_test_df = pd.DataFrame(X_test)\n",
    "y_test_df = pd.DataFrame(y_test)\n",
    "\n",
    "\n",
    "# combine X_test_df and y_test_df into a single DataFrame for visualization\n",
    "test_data = pd.concat([X_test_df, y_test_df], axis=1)\n",
    "\n",
    "# display pairplot\n",
    "print('Plot of joint distributions of each of our features along with kernel density estimation plots')\n",
    "sns.pairplot(test_data, diag_kind='kde')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "h_s58iVS4ooy"
   },
   "source": [
    "Print the a list of summary statistics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "7wZrDmZR4qUj",
    "outputId": "4783a486-0823-43f8-a7f2-ccfb2050461b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: tabulate in /usr/local/lib/python3.10/dist-packages (0.9.0)\n",
      "+--------------------+-------------+------------+------------+\n",
      "| Statistic          |    Column 0 |   Column 1 |   Column 2 |\n",
      "+====================+=============+============+============+\n",
      "| Mean               | -0.00987319 | -0.0145918 |  0.0165762 |\n",
      "+--------------------+-------------+------------+------------+\n",
      "| Standard Deviation |  1.00235    |  0.993221  |  0.99857   |\n",
      "+--------------------+-------------+------------+------------+\n",
      "| Minimum Value      | -1.93631    | -0.997813  | -1.47888   |\n",
      "+--------------------+-------------+------------+------------+\n",
      "| Maximum Value      |  3.79625    |  3.18743   |  1.28862   |\n",
      "+--------------------+-------------+------------+------------+\n"
     ]
    }
   ],
   "source": [
    "!pip install tabulate\n",
    "from tabulate import tabulate\n",
    "\n",
    "mean = np.mean(X_train, axis=0)\n",
    "std_dev = np.std(X_train, axis=0)\n",
    "min_value = np.min(X_train, axis=0)\n",
    "max_value = np.max(X_train, axis=0)\n",
    "\n",
    "\n",
    "summary_data = [\n",
    "    ['Mean', *mean],\n",
    "    ['Standard Deviation', *std_dev],\n",
    "    ['Minimum Value', *min_value],\n",
    "    ['Maximum Value', *max_value]\n",
    "]\n",
    "\n",
    "# print the a list of summary statistics\n",
    "table = tabulate(summary_data, headers=['Statistic'] + [f'Column {i}' for i in range(X_train.shape[1])], tablefmt='grid')\n",
    "print(table)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "eoc1lfFq5lSm"
   },
   "source": [
    "Normalize the features so every value will fall between 0 and 1 using MinMaxScaler."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ksMRPZOv51xM",
    "outputId": "cf720088-da75-4f28-8ffc-d6f908040796"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Normalized Train Data Summary:\n",
      "+----------------------------+------------+------------+------------+\n",
      "| Statistic                  |   Column 0 |   Column 1 |   Column 2 |\n",
      "+============================+============+============+============+\n",
      "| Mean (Train)               |   0.338664 |   0.241464 |   0.539993 |\n",
      "+----------------------------+------------+------------+------------+\n",
      "| Standard Deviation (Train) |   0.174343 |   0.239475 |   0.359484 |\n",
      "+----------------------------+------------+------------+------------+\n",
      "| Minimum Value (Train)      |   0        |   0        |   0        |\n",
      "+----------------------------+------------+------------+------------+\n",
      "| Maximum Value (Train)      |   1        |   1        |   1        |\n",
      "+----------------------------+------------+------------+------------+\n",
      "\n",
      "Normalized Test Data Summary:\n",
      "+---------------------------+------------+------------+------------+\n",
      "| Statistic                 |   Column 0 |   Column 1 |   Column 2 |\n",
      "+===========================+============+============+============+\n",
      "| Mean (Test)               |  0.344663  | 0.252358   |   0.510417 |\n",
      "+---------------------------+------------+------------+------------+\n",
      "| Standard Deviation (Test) |  0.172618  | 0.24481    |   0.362408 |\n",
      "+---------------------------+------------+------------+------------+\n",
      "| Minimum Value (Test)      |  0.0227273 | 0.00119048 |   0        |\n",
      "+---------------------------+------------+------------+------------+\n",
      "| Maximum Value (Test)      |  0.772727  | 0.999603   |   1        |\n",
      "+---------------------------+------------+------------+------------+\n"
     ]
    }
   ],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "scaler = MinMaxScaler()\n",
    "\n",
    "# fit the scaler to your training data and transform it\n",
    "X_train_normalized = scaler.fit_transform(X_train)\n",
    "\n",
    "# transform the testing data using the same scaler\n",
    "X_test_normalized = scaler.transform(X_test)\n",
    "\n",
    "# print the normalized data summary for both train and test sets\n",
    "print(\"Normalized Train Data Summary:\")\n",
    "normalized_train_mean = np.mean(X_train_normalized, axis=0)\n",
    "normalized_train_std_dev = np.std(X_train_normalized, axis=0)\n",
    "normalized_train_min_value = np.min(X_train_normalized, axis=0)\n",
    "normalized_train_max_value = np.max(X_train_normalized, axis=0)\n",
    "\n",
    "normalized_train_summary_data = [\n",
    "    ['Mean (Train)', *normalized_train_mean],\n",
    "    ['Standard Deviation (Train)', *normalized_train_std_dev],\n",
    "    ['Minimum Value (Train)', *normalized_train_min_value],\n",
    "    ['Maximum Value (Train)', *normalized_train_max_value]\n",
    "]\n",
    "\n",
    "print(tabulate(normalized_train_summary_data, headers=['Statistic'] + [f'Column {i}' for i in range(X_train.shape[1])], tablefmt='grid'))\n",
    "\n",
    "print(\"\\nNormalized Test Data Summary:\")\n",
    "normalized_test_mean = np.mean(X_test_normalized, axis=0)\n",
    "normalized_test_std_dev = np.std(X_test_normalized, axis=0)\n",
    "normalized_test_min_value = np.min(X_test_normalized, axis=0)\n",
    "normalized_test_max_value = np.max(X_test_normalized, axis=0)\n",
    "\n",
    "normalized_test_summary_data = [\n",
    "    ['Mean (Test)', *normalized_test_mean],\n",
    "    ['Standard Deviation (Test)', *normalized_test_std_dev],\n",
    "    ['Minimum Value (Test)', *normalized_test_min_value],\n",
    "    ['Maximum Value (Test)', *normalized_test_max_value]\n",
    "]\n",
    "\n",
    "print(tabulate(normalized_test_summary_data, headers=['Statistic'] + [f'Column {i}' for i in range(X_test.shape[1])], tablefmt='grid'))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "23Ayw69L6Ba0"
   },
   "source": [
    "<h2>5. Developing a model that does better than a baseline</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "yqohJMA66CRi"
   },
   "source": [
    "<h3>5.1 Build and Train a TensorFlow Model</h3>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "j68usgAu6nfH"
   },
   "source": [
    "<h4>5.1.1 Build the model: Linear regression with multiple inputs</h4>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "r6AKQ6oE6ErM"
   },
   "source": [
    "Define the neural network architecture using TensorFlow's Keras API:\n",
    "\n",
    "\n",
    "* choose the number of layers\n",
    "* choose number of neurons in each layer\n",
    "* choose activation functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "IWIpz7It6IF7",
    "outputId": "06d7bd15-9951-4719-b501-e9093eb92080"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X shape: (2240, 3)\n",
      "y shape: (2240,)\n"
     ]
    }
   ],
   "source": [
    "print(\"X shape:\", X.shape)\n",
    "print(\"y shape:\", y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "TI1Yqyfb6K0m",
    "outputId": "07ffb300-e9b5-4ffe-dee4-3de9b35df282"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/70\n",
      "45/45 [==============================] - 2s 7ms/step - loss: 223494.9219 - mean_absolute_error: 308.2108 - mean_squared_error: 223494.9219 - val_loss: 11001.8799 - val_mean_absolute_error: 41.8263 - val_mean_squared_error: 11001.8799\n",
      "Epoch 2/70\n",
      "45/45 [==============================] - 0s 4ms/step - loss: 221097.1406 - mean_absolute_error: 306.0172 - mean_squared_error: 221097.1406 - val_loss: 10594.1133 - val_mean_absolute_error: 39.0780 - val_mean_squared_error: 10594.1133\n",
      "Epoch 3/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 214883.2344 - mean_absolute_error: 300.1418 - mean_squared_error: 214883.2344 - val_loss: 9719.2783 - val_mean_absolute_error: 33.5240 - val_mean_squared_error: 9719.2783\n",
      "Epoch 4/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 202256.9531 - mean_absolute_error: 288.8934 - mean_squared_error: 202256.9531 - val_loss: 8604.3271 - val_mean_absolute_error: 33.4324 - val_mean_squared_error: 8604.3271\n",
      "Epoch 5/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 181613.7188 - mean_absolute_error: 272.1910 - mean_squared_error: 181613.7188 - val_loss: 8321.5479 - val_mean_absolute_error: 50.0140 - val_mean_squared_error: 8321.5479\n",
      "Epoch 6/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 154602.5781 - mean_absolute_error: 253.1829 - mean_squared_error: 154602.5781 - val_loss: 11193.0449 - val_mean_absolute_error: 84.4184 - val_mean_squared_error: 11193.0449\n",
      "Epoch 7/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 125668.9375 - mean_absolute_error: 235.7440 - mean_squared_error: 125668.9375 - val_loss: 19252.3164 - val_mean_absolute_error: 127.8171 - val_mean_squared_error: 19252.3164\n",
      "Epoch 8/70\n",
      "45/45 [==============================] - 0s 4ms/step - loss: 102320.5781 - mean_absolute_error: 223.8721 - mean_squared_error: 102320.5781 - val_loss: 32061.0781 - val_mean_absolute_error: 172.1080 - val_mean_squared_error: 32061.0781\n",
      "Epoch 9/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 88521.6797 - mean_absolute_error: 221.8618 - mean_squared_error: 88521.6797 - val_loss: 44144.4883 - val_mean_absolute_error: 203.4879 - val_mean_squared_error: 44144.4883\n",
      "Epoch 10/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 81767.5391 - mean_absolute_error: 221.5334 - mean_squared_error: 81767.5391 - val_loss: 50316.8359 - val_mean_absolute_error: 216.4044 - val_mean_squared_error: 50316.8359\n",
      "Epoch 11/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 77405.5547 - mean_absolute_error: 218.9122 - mean_squared_error: 77405.5547 - val_loss: 52701.6992 - val_mean_absolute_error: 219.6611 - val_mean_squared_error: 52701.6992\n",
      "Epoch 12/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 73594.2500 - mean_absolute_error: 214.0762 - mean_squared_error: 73594.2500 - val_loss: 51156.9180 - val_mean_absolute_error: 213.8429 - val_mean_squared_error: 51156.9180\n",
      "Epoch 13/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 69960.9609 - mean_absolute_error: 207.3250 - mean_squared_error: 69960.9609 - val_loss: 49644.4023 - val_mean_absolute_error: 207.5030 - val_mean_squared_error: 49644.4023\n",
      "Epoch 14/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 66238.6406 - mean_absolute_error: 200.4871 - mean_squared_error: 66238.6406 - val_loss: 48897.9961 - val_mean_absolute_error: 202.1909 - val_mean_squared_error: 48897.9961\n",
      "Epoch 15/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 62433.6133 - mean_absolute_error: 193.3019 - mean_squared_error: 62433.6133 - val_loss: 46530.3906 - val_mean_absolute_error: 192.1940 - val_mean_squared_error: 46530.3906\n",
      "Epoch 16/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 58716.0547 - mean_absolute_error: 185.6383 - mean_squared_error: 58716.0547 - val_loss: 45314.6055 - val_mean_absolute_error: 184.2363 - val_mean_squared_error: 45314.6055\n",
      "Epoch 17/70\n",
      "45/45 [==============================] - 0s 4ms/step - loss: 54890.4453 - mean_absolute_error: 176.5581 - mean_squared_error: 54890.4453 - val_loss: 42301.3164 - val_mean_absolute_error: 172.2072 - val_mean_squared_error: 42301.3164\n",
      "Epoch 18/70\n",
      "45/45 [==============================] - 0s 4ms/step - loss: 51248.3789 - mean_absolute_error: 169.6262 - mean_squared_error: 51248.3789 - val_loss: 38584.1094 - val_mean_absolute_error: 163.7945 - val_mean_squared_error: 38584.1094\n",
      "Epoch 19/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 47810.5859 - mean_absolute_error: 161.2748 - mean_squared_error: 47810.5859 - val_loss: 37411.7539 - val_mean_absolute_error: 162.0217 - val_mean_squared_error: 37411.7539\n",
      "Epoch 20/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 44606.1211 - mean_absolute_error: 155.7942 - mean_squared_error: 44606.1211 - val_loss: 33832.1328 - val_mean_absolute_error: 153.8052 - val_mean_squared_error: 33832.1328\n",
      "Epoch 21/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 41524.8594 - mean_absolute_error: 148.8961 - mean_squared_error: 41524.8594 - val_loss: 33735.2266 - val_mean_absolute_error: 152.6777 - val_mean_squared_error: 33735.2266\n",
      "Epoch 22/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 38833.7109 - mean_absolute_error: 142.0500 - mean_squared_error: 38833.7109 - val_loss: 33597.6758 - val_mean_absolute_error: 150.7929 - val_mean_squared_error: 33597.6758\n",
      "Epoch 23/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 36488.3633 - mean_absolute_error: 138.9795 - mean_squared_error: 36488.3633 - val_loss: 32028.6621 - val_mean_absolute_error: 146.6145 - val_mean_squared_error: 32028.6621\n",
      "Epoch 24/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 34406.8828 - mean_absolute_error: 135.0432 - mean_squared_error: 34406.8828 - val_loss: 31391.1777 - val_mean_absolute_error: 148.5199 - val_mean_squared_error: 31391.1777\n",
      "Epoch 25/70\n",
      "45/45 [==============================] - 0s 4ms/step - loss: 32709.0410 - mean_absolute_error: 132.0249 - mean_squared_error: 32709.0410 - val_loss: 31625.0117 - val_mean_absolute_error: 151.6462 - val_mean_squared_error: 31625.0117\n",
      "Epoch 26/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 31152.3184 - mean_absolute_error: 129.6079 - mean_squared_error: 31152.3184 - val_loss: 30717.5664 - val_mean_absolute_error: 152.0321 - val_mean_squared_error: 30717.5664\n",
      "Epoch 27/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 29801.5820 - mean_absolute_error: 127.7589 - mean_squared_error: 29801.5820 - val_loss: 31605.4844 - val_mean_absolute_error: 155.3038 - val_mean_squared_error: 31605.4844\n",
      "Epoch 28/70\n",
      "45/45 [==============================] - 0s 4ms/step - loss: 28677.2988 - mean_absolute_error: 127.3160 - mean_squared_error: 28677.2988 - val_loss: 30006.2637 - val_mean_absolute_error: 153.2270 - val_mean_squared_error: 30006.2637\n",
      "Epoch 29/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 27651.6094 - mean_absolute_error: 125.2508 - mean_squared_error: 27651.6094 - val_loss: 29600.7363 - val_mean_absolute_error: 152.7391 - val_mean_squared_error: 29600.7363\n",
      "Epoch 30/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 26798.8145 - mean_absolute_error: 123.9868 - mean_squared_error: 26798.8145 - val_loss: 29301.0371 - val_mean_absolute_error: 152.2370 - val_mean_squared_error: 29301.0371\n",
      "Epoch 31/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 26078.4258 - mean_absolute_error: 122.4903 - mean_squared_error: 26078.4258 - val_loss: 29553.1680 - val_mean_absolute_error: 153.0167 - val_mean_squared_error: 29553.1680\n",
      "Epoch 32/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 25613.1777 - mean_absolute_error: 121.8204 - mean_squared_error: 25613.1777 - val_loss: 29332.7715 - val_mean_absolute_error: 152.6248 - val_mean_squared_error: 29332.7715\n",
      "Epoch 33/70\n",
      "45/45 [==============================] - 0s 4ms/step - loss: 24977.9590 - mean_absolute_error: 120.4702 - mean_squared_error: 24977.9590 - val_loss: 28440.3984 - val_mean_absolute_error: 150.3515 - val_mean_squared_error: 28440.3984\n",
      "Epoch 34/70\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 24387.9863 - mean_absolute_error: 119.1245 - mean_squared_error: 24387.9863 - val_loss: 28285.1094 - val_mean_absolute_error: 150.1286 - val_mean_squared_error: 28285.1094\n",
      "Epoch 35/70\n",
      "45/45 [==============================] - 0s 5ms/step - loss: 23823.7832 - mean_absolute_error: 117.9632 - mean_squared_error: 23823.7832 - val_loss: 26612.8320 - val_mean_absolute_error: 145.1003 - val_mean_squared_error: 26612.8320\n",
      "Epoch 36/70\n",
      "45/45 [==============================] - 0s 4ms/step - loss: 23331.0879 - mean_absolute_error: 116.3050 - mean_squared_error: 23331.0879 - val_loss: 26635.3379 - val_mean_absolute_error: 145.4595 - val_mean_squared_error: 26635.3379\n",
      "Epoch 37/70\n",
      "45/45 [==============================] - 0s 5ms/step - loss: 22847.6504 - mean_absolute_error: 115.3476 - mean_squared_error: 22847.6504 - val_loss: 25543.9434 - val_mean_absolute_error: 141.7674 - val_mean_squared_error: 25543.9434\n",
      "Epoch 38/70\n",
      "45/45 [==============================] - 0s 5ms/step - loss: 22411.9590 - mean_absolute_error: 113.7700 - mean_squared_error: 22411.9590 - val_loss: 25554.8359 - val_mean_absolute_error: 142.4314 - val_mean_squared_error: 25554.8359\n",
      "Epoch 39/70\n",
      "45/45 [==============================] - 0s 5ms/step - loss: 21955.5137 - mean_absolute_error: 113.0134 - mean_squared_error: 21955.5137 - val_loss: 24463.9727 - val_mean_absolute_error: 138.5058 - val_mean_squared_error: 24463.9727\n",
      "Epoch 40/70\n",
      "45/45 [==============================] - 0s 5ms/step - loss: 21550.6035 - mean_absolute_error: 111.5689 - mean_squared_error: 21550.6035 - val_loss: 24164.8574 - val_mean_absolute_error: 137.7410 - val_mean_squared_error: 24164.8574\n",
      "Epoch 41/70\n",
      "45/45 [==============================] - 0s 5ms/step - loss: 21071.7012 - mean_absolute_error: 110.4598 - mean_squared_error: 21071.7012 - val_loss: 23392.3594 - val_mean_absolute_error: 135.1088 - val_mean_squared_error: 23392.3594\n",
      "Epoch 42/70\n",
      "45/45 [==============================] - 0s 4ms/step - loss: 20591.2168 - mean_absolute_error: 108.9687 - mean_squared_error: 20591.2168 - val_loss: 22957.0312 - val_mean_absolute_error: 133.8921 - val_mean_squared_error: 22957.0312\n",
      "Epoch 43/70\n",
      "45/45 [==============================] - 0s 5ms/step - loss: 20092.3008 - mean_absolute_error: 107.6873 - mean_squared_error: 20092.3008 - val_loss: 22045.1719 - val_mean_absolute_error: 130.5406 - val_mean_squared_error: 22045.1719\n",
      "Epoch 44/70\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 19604.6445 - mean_absolute_error: 106.5875 - mean_squared_error: 19604.6445 - val_loss: 20494.3320 - val_mean_absolute_error: 124.3833 - val_mean_squared_error: 20494.3320\n",
      "Epoch 45/70\n",
      "45/45 [==============================] - 0s 5ms/step - loss: 19000.7070 - mean_absolute_error: 104.5900 - mean_squared_error: 19000.7070 - val_loss: 20259.8340 - val_mean_absolute_error: 123.9081 - val_mean_squared_error: 20259.8340\n",
      "Epoch 46/70\n",
      "45/45 [==============================] - 0s 4ms/step - loss: 18427.8652 - mean_absolute_error: 102.6986 - mean_squared_error: 18427.8652 - val_loss: 18888.2363 - val_mean_absolute_error: 118.5550 - val_mean_squared_error: 18888.2363\n",
      "Epoch 47/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 17852.5996 - mean_absolute_error: 101.0767 - mean_squared_error: 17852.5996 - val_loss: 18346.8164 - val_mean_absolute_error: 116.4748 - val_mean_squared_error: 18346.8164\n",
      "Epoch 48/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 17196.5879 - mean_absolute_error: 99.3318 - mean_squared_error: 17196.5879 - val_loss: 16578.2266 - val_mean_absolute_error: 109.4669 - val_mean_squared_error: 16578.2266\n",
      "Epoch 49/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 16575.5469 - mean_absolute_error: 97.3024 - mean_squared_error: 16575.5469 - val_loss: 15294.2773 - val_mean_absolute_error: 105.3722 - val_mean_squared_error: 15294.2773\n",
      "Epoch 50/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 15954.5859 - mean_absolute_error: 96.0697 - mean_squared_error: 15954.5859 - val_loss: 14742.5029 - val_mean_absolute_error: 102.3964 - val_mean_squared_error: 14742.5029\n",
      "Epoch 51/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 15369.2334 - mean_absolute_error: 94.5186 - mean_squared_error: 15369.2334 - val_loss: 13765.6113 - val_mean_absolute_error: 98.7980 - val_mean_squared_error: 13765.6113\n",
      "Epoch 52/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 14832.1973 - mean_absolute_error: 93.0797 - mean_squared_error: 14832.1973 - val_loss: 13465.4248 - val_mean_absolute_error: 95.9608 - val_mean_squared_error: 13465.4248\n",
      "Epoch 53/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 14276.2314 - mean_absolute_error: 90.9524 - mean_squared_error: 14276.2314 - val_loss: 12124.4316 - val_mean_absolute_error: 92.3944 - val_mean_squared_error: 12124.4316\n",
      "Epoch 54/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 13729.0801 - mean_absolute_error: 89.2608 - mean_squared_error: 13729.0801 - val_loss: 10989.3750 - val_mean_absolute_error: 89.2417 - val_mean_squared_error: 10989.3750\n",
      "Epoch 55/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 13257.8438 - mean_absolute_error: 89.2166 - mean_squared_error: 13257.8438 - val_loss: 11010.1387 - val_mean_absolute_error: 86.3633 - val_mean_squared_error: 11010.1387\n",
      "Epoch 56/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 12770.7930 - mean_absolute_error: 87.3134 - mean_squared_error: 12770.7930 - val_loss: 10171.4326 - val_mean_absolute_error: 83.3231 - val_mean_squared_error: 10171.4326\n",
      "Epoch 57/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 12341.0332 - mean_absolute_error: 85.4665 - mean_squared_error: 12341.0332 - val_loss: 9810.3896 - val_mean_absolute_error: 80.5459 - val_mean_squared_error: 9810.3896\n",
      "Epoch 58/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 11935.9150 - mean_absolute_error: 84.4266 - mean_squared_error: 11935.9150 - val_loss: 9460.2676 - val_mean_absolute_error: 77.7892 - val_mean_squared_error: 9460.2676\n",
      "Epoch 59/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 11565.8154 - mean_absolute_error: 82.7788 - mean_squared_error: 11565.8154 - val_loss: 8434.8682 - val_mean_absolute_error: 75.1477 - val_mean_squared_error: 8434.8682\n",
      "Epoch 60/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 11174.0645 - mean_absolute_error: 80.9148 - mean_squared_error: 11174.0645 - val_loss: 8020.5312 - val_mean_absolute_error: 72.2481 - val_mean_squared_error: 8020.5312\n",
      "Epoch 61/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 10790.9404 - mean_absolute_error: 79.7977 - mean_squared_error: 10790.9404 - val_loss: 7932.9873 - val_mean_absolute_error: 69.4802 - val_mean_squared_error: 7932.9873\n",
      "Epoch 62/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 10513.6807 - mean_absolute_error: 78.1474 - mean_squared_error: 10513.6807 - val_loss: 7492.5879 - val_mean_absolute_error: 66.9644 - val_mean_squared_error: 7492.5879\n",
      "Epoch 63/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 10212.3467 - mean_absolute_error: 76.4558 - mean_squared_error: 10212.3467 - val_loss: 6946.0703 - val_mean_absolute_error: 64.6037 - val_mean_squared_error: 6946.0703\n",
      "Epoch 64/70\n",
      "45/45 [==============================] - 0s 4ms/step - loss: 10028.8525 - mean_absolute_error: 76.0070 - mean_squared_error: 10028.8525 - val_loss: 7106.5801 - val_mean_absolute_error: 61.9902 - val_mean_squared_error: 7106.5801\n",
      "Epoch 65/70\n",
      "45/45 [==============================] - 0s 4ms/step - loss: 9695.3291 - mean_absolute_error: 74.1522 - mean_squared_error: 9695.3291 - val_loss: 6665.9434 - val_mean_absolute_error: 59.6531 - val_mean_squared_error: 6665.9434\n",
      "Epoch 66/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 9521.7285 - mean_absolute_error: 72.9964 - mean_squared_error: 9521.7285 - val_loss: 6481.3779 - val_mean_absolute_error: 57.3602 - val_mean_squared_error: 6481.3779\n",
      "Epoch 67/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 9283.5605 - mean_absolute_error: 72.3780 - mean_squared_error: 9283.5605 - val_loss: 6616.4365 - val_mean_absolute_error: 56.9530 - val_mean_squared_error: 6616.4365\n",
      "Epoch 68/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 9094.0859 - mean_absolute_error: 71.1278 - mean_squared_error: 9094.0859 - val_loss: 6245.7173 - val_mean_absolute_error: 56.0379 - val_mean_squared_error: 6245.7173\n",
      "Epoch 69/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 8915.0967 - mean_absolute_error: 70.8430 - mean_squared_error: 8915.0967 - val_loss: 5947.3608 - val_mean_absolute_error: 55.3960 - val_mean_squared_error: 5947.3608\n",
      "Epoch 70/70\n",
      "45/45 [==============================] - 0s 3ms/step - loss: 8756.2188 - mean_absolute_error: 70.8947 - mean_squared_error: 8756.2188 - val_loss: 6107.5942 - val_mean_absolute_error: 55.9083 - val_mean_squared_error: 6107.5942\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "\n",
    "my_model = tf.keras.Sequential([\n",
    "    tf.keras.layers.Dense(64, activation='relu', input_shape=(X_train.shape[1],)),\n",
    "    tf.keras.layers.Dense(32, activation='relu'),\n",
    "    tf.keras.layers.Dense(1)\n",
    "])\n",
    "\n",
    "my_model.compile(loss='mean_squared_error', optimizer='adam', metrics=['mean_absolute_error','mean_squared_error'])\n",
    "\n",
    "history = my_model.fit(X_train_normalized, y_train, epochs=70, batch_size=32, validation_split=0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2Ie2Ds9k7Z90"
   },
   "source": [
    "<h4> 5.1.2 Inspect the model</h4>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 875
    },
    "id": "_c7qjsz77fgw",
    "outputId": "03ddd5fa-0c25-4336-b6ca-8b895d779e8d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_5\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense_21 (Dense)            (None, 64)                256       \n",
      "                                                                 \n",
      " dense_22 (Dense)            (None, 32)                2080      \n",
      "                                                                 \n",
      " dense_23 (Dense)            (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 2369 (9.25 KB)\n",
      "Trainable params: 2369 (9.25 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABKUAAAJOCAYAAABm7rQwAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAADkmUlEQVR4nOzdd3RU1frG8e9Meg8B0iCE3quhBaQoSChSBK+CBVB+YgEVsTfEdu0KCha8KqhgQQUVFQhFUDpIb9JrChCSkITUOb8/hhmJBAiQ5KQ8n7VmZcrOOc8Mdznnvtn73RbDMAxERERERERERERKkNXsACIiIiIiIiIiUvGoKCUiIiIiIiIiIiVORSkRERERERERESlxKkqJiIiIiIiIiEiJU1FKRERERERERERKnIpSIiIiIiIiIiJS4lSUEhERERERERGREqeilIiIiIiIiIiIlDgVpUREREREREREpMSpKCUiUkgWi4Xx48df8u/t378fi8XC1KlTizyTiIiISFEaPnw4vr6+JXpOXSuJVFwqSolImTJ16lQsFgsWi4U///zznNcNwyAiIgKLxcL1119vQsLL9/vvv2OxWPjuu+/MjiIiIlLhlcdrjry8PMLDw7FYLPz2229mxykSGRkZjB8/nt9//73Ij+349y/ods899xT5+UQqIlezA4iIXA5PT09mzJjB1Vdfne/5JUuWcPjwYTw8PExKJiIiIuVJebrmWLRoEXFxcdSsWZPp06fTq1cvsyNdsYyMDJ5//nkAunbtWuTHv+666xg6dOg5z9evX7/IzyVSEakoJSJlUu/evZk5cybvvvsurq7//KdsxowZREVFcfz4cRPTiYiISHlRnq45vvzyS6666iqGDRvGU089RXp6Oj4+PmbHKtXq16/Pbbfddsm/l5GRgbe39znP5+bmYrPZcHd3v+xM+neT8kTL90SkTBoyZAgnTpwgNjbW+Vx2djbfffcdt9xyS4G/k56ezsMPP0xERAQeHh40aNCAN998E8Mw8o3LysrioYceomrVqvj5+dGvXz8OHz5c4DGPHDnCnXfeSUhICB4eHjRp0oRPP/206N5oAfbu3ct//vMfgoKC8Pb2pn379vzyyy/njHvvvfdo0qQJ3t7eVKpUidatWzNjxgzn66dOnWLMmDHUrFkTDw8PgoODue666/jrr7+KNb+IiEhZcjnXHDabjQkTJtCkSRM8PT0JCQnh7rvv5uTJk/nG/fjjj/Tp04fw8HA8PDyoU6cOL774Inl5efnGde3alaZNm7Jt2zauueYavL29qVatGq+//nqh38fp06eZNWsWgwcP5qabbuL06dP8+OOP5x2/d+9eYmJi8PHxITw8nBdeeOGca6avv/6aqKgo/Pz88Pf3p1mzZkycOPGc4xTmuuXfunbtWuDMp+HDh1OzZk3A3ouqatWqADz//PPOpXVn9wDdsWMHN954I0FBQXh6etK6dWt++umni57/Ujj+fdatW0fnzp3x9vbmqaeecvbKevPNN5kwYQJ16tTBw8ODbdu2AfaZa506dcLHx4fAwED69+/P9u3b8x17/PjxWCwWtm3bxi233EKlSpXOmbUnUpapKCUiZVLNmjWJjo7mq6++cj7322+/kZKSwuDBg88ZbxgG/fr145133qFnz568/fbbNGjQgEcffZSxY8fmG/t///d/TJgwgR49evDqq6/i5uZGnz59zjlmQkIC7du3Z8GCBYwePZqJEydSt25dRowYwYQJE4r8PTvO2aFDB+bNm8d9993Hyy+/TGZmJv369WPWrFnOcR9//DEPPPAAjRs3ZsKECTz//PO0bNmSVatWOcfcc889fPDBBwwaNIj333+fRx55BC8vr3MuhkRERCqyS73mALj77rt59NFH6dixIxMnTuSOO+5g+vTpxMTEkJOT4xw3depUfH19GTt2LBMnTiQqKopx48bxxBNPnHPMkydP0rNnT1q0aMFbb71Fw4YNefzxxwvdG+qnn34iLS2NwYMHExoaSteuXZk+fXqBY/Py8ujZsychISG8/vrrREVF8dxzz/Hcc885x8TGxjJkyBAqVarEa6+9xquvvkrXrl1ZtmyZc0xhr1suV9WqVfnggw8AuOGGG/jiiy/44osvGDhwIABbt26lffv2bN++nSeeeIK33noLHx8fBgwYUOjzZ2Zmcvz48XNu2dnZ+cadOHGCXr160bJlSyZMmMA111zjfO2zzz7jvffeY+TIkbz11lsEBQWxYMECYmJiSExMZPz48YwdO5bly5fTsWNH9u/ff06O//znP2RkZPDf//6Xu+666zI/MZFSyBARKUM+++wzAzDWrFljTJo0yfDz8zMyMjIMwzCM//znP8Y111xjGIZhREZGGn369HH+3uzZsw3AeOmll/Id78YbbzQsFouxe/duwzAMY8OGDQZg3HffffnG3XLLLQZgPPfcc87nRowYYYSFhRnHjx/PN3bw4MFGQECAM9e+ffsMwPjss88u+N4WL15sAMbMmTPPO2bMmDEGYPzxxx/O506dOmXUqlXLqFmzppGXl2cYhmH079/faNKkyQXPFxAQYIwaNeqCY0RERCqqy73m+OOPPwzAmD59er7jzZ0795znHcc729133214e3sbmZmZzue6dOliAMbnn3/ufC4rK8sIDQ01Bg0aVKj3c/311xsdO3Z0Pp4yZYrh6upqJCYm5hs3bNgwAzDuv/9+53M2m83o06eP4e7ubhw7dswwDMN48MEHDX9/fyM3N/e85yzsdUtB10pdunQxunTpcs4xhw0bZkRGRjofHzt27JxrNIdu3boZzZo1y/dZ2mw2o0OHDka9evXOm9sBOO/tq6++ypcVMD788MN8v+94X/7+/ud8zi1btjSCg4ONEydOOJ/buHGjYbVajaFDhzqfe+655wzAGDJkyEXzipRFmiklImWWY+r5nDlzOHXqFHPmzDnvNPpff/0VFxcXHnjggXzPP/zwwxiG4fwr46+//gpwzrgxY8bke2wYBt9//z19+/bFMIx8fzmLiYkhJSWlWJbB/frrr7Rt2zbftG1fX19GjhzJ/v37ndPBAwMDOXz4MGvWrDnvsQIDA1m1ahVHjx4t8pwiIiLlyaVcc8ycOZOAgACuu+66fNcHUVFR+Pr6snjxYudYLy8v5/1Tp05x/PhxOnXqREZGBjt27Mh3XF9f33y9jdzd3Wnbti179+69aP4TJ04wb948hgwZ4nxu0KBBWCwWvv322wJ/Z/To0c77FouF0aNHk52dzYIFCwD7dUR6enq+ZY3/VtjrluKQlJTEokWLuOmmm5yf7fHjxzlx4gQxMTHs2rWLI0eOXPQ4/fv3JzY29pzb2TOhADw8PLjjjjsKPMagQYOcywwB4uLi2LBhA8OHDycoKMj5fPPmzbnuuuuc16Nn025/Ul6p0bmIlFlVq1ale/fuzJgxg4yMDPLy8rjxxhsLHHvgwAHCw8Px8/PL93yjRo2crzt+Wq1W6tSpk29cgwYN8j0+duwYycnJTJkyhSlTphR4zsTExMt6Xxdy4MAB2rVrd87zZ7+Ppk2b8vjjj7NgwQLatm1L3bp16dGjB7fccgsdO3Z0/s7rr7/OsGHDiIiIICoqit69ezN06FBq165d5LlFRETKsku55ti1axcpKSkEBwcX+PrZ1wdbt27lmWeeYdGiRaSmpuYbl5KSku9x9erVsVgs+Z6rVKkSmzZtumj+b775hpycHFq1asXu3budz7dr147p06czatSofOOtVus51wOO3eYcS8vuu+8+vv32W3r16kW1atXo0aMHN910Ez179nT+TmGvW4rD7t27MQyDZ599lmeffbbAMYmJiVSrVu2Cx6levTrdu3e/6PmqVat23ubltWrVyvfYcd357+tLsH828+bNO6eZ+b+PIVJeqCglImXaLbfcwl133UV8fDy9evUiMDCwRM5rs9kAuO222xg2bFiBY5o3b14iWQrSqFEjdu7cyZw5c5g7dy7ff/8977//PuPGjXNum3zTTTfRqVMnZs2axfz583njjTd47bXX+OGHH8rFFtEiIiJFqbDXHDabjeDg4PP2a3LMmElOTqZLly74+/vzwgsvUKdOHTw9Pfnrr794/PHHndcaDi4uLgUez/hX8/GCOLKc/ceps+3du/eS/ygVHBzMhg0bmDdvHr/99hu//fYbn332GUOHDmXatGmXdKyCWCyWAt/bv5vAn4/j83vkkUeIiYkpcEzdunUvP+C/nD3r7VJeK4rji5RlKkqJSJl2ww03cPfdd7Ny5Uq++eab846LjIxkwYIFnDp1Kt9sKcfU+MjISOdPm83Gnj178v31aufOnfmO59iZLy8vr1B/PSsqkZGR52SBc98HgI+PDzfffDM333wz2dnZDBw4kJdffpknn3wST09PAMLCwrjvvvu47777SExM5KqrruLll19WUUpERORfCnvNUadOHRYsWEDHjh0vWEj4/fffOXHiBD/88AOdO3d2Pr9v374izb1v3z6WL1/O6NGj6dKlS77XbDYbt99+OzNmzOCZZ57J9/zevXuds6MA/v77bwDnzndgX0LYt29f+vbti81m47777uOjjz7i2WefpW7dupd03fJvlSpVKnBpomOWkcO/Z485OIpsbm5uJXqtVhiO932+z6ZKlSr5ZkmJlGfqKSUiZZqvry8ffPAB48ePp2/fvucd17t3b/Ly8pg0aVK+59955x0sFouzCOP4+e677+Yb9+/d9FxcXBg0aBDff/89W7ZsOed8x44du5y3c1G9e/dm9erVrFixwvlceno6U6ZMoWbNmjRu3Biw9444m7u7O40bN8YwDHJycsjLyztnWUBwcDDh4eFkZWUVS3YREZGyrLDXHDfddBN5eXm8+OKL57yWm5tLcnIy8M/Mp7NnA2VnZ/P+++8XaW7HLKnHHnuMG2+8Md/tpptuokuXLgXO6jr7mskwDCZNmoSbmxvdunUDzr3WsFqtzlnijmuJwl63FKROnTrs2LEj3zXVxo0b8+3uB+Dt7Q3g/FwdgoOD6dq1Kx999BFxcXHnHL+4rtUKIywsjJYtWzJt2rR8ubds2cL8+fPp3bu3adlESppmSolImXe+5XNn69u3L9dccw1PP/00+/fvp0WLFsyfP58ff/yRMWPGOHtItWzZkiFDhvD++++TkpJChw4dWLhwYb7+Cw6vvvoqixcvpl27dtx11100btyYpKQk/vrrLxYsWEBSUtJlvZ/vv//+nOamjvf5xBNP8NVXX9GrVy8eeOABgoKCmDZtGvv27eP777/HarX/raFHjx6EhobSsWNHQkJC2L59O5MmTaJPnz74+fmRnJxM9erVufHGG2nRogW+vr4sWLCANWvW8NZbb11WbhERkfKuMNccXbp04e677+aVV15hw4YN9OjRAzc3N3bt2sXMmTOZOHEiN954Ix06dKBSpUoMGzaMBx54AIvFwhdffFGo5XiXYvr06bRs2ZKIiIgCX+/Xrx/3338/f/31F1dddRUAnp6ezJ07l2HDhtGuXTt+++03fvnlF5566inn8sP/+7//IykpiWuvvZbq1atz4MAB3nvvPVq2bOnsGVXY65aC3Hnnnbz99tvExMQwYsQIEhMT+fDDD2nSpEm+/lteXl40btyYb775hvr16xMUFETTpk1p2rQpkydP5uqrr6ZZs2bcdddd1K5dm4SEBFasWMHhw4fZuHHjRT+/v//+my+//PKc50NCQrjuuusu+vvn88Ybb9CrVy+io6MZMWIEp0+f5r333iMgIIDx48df9nFFyhyztv0TEbkcZ2/PfCH/3p7ZMOxbED/00ENGeHi44ebmZtSrV8944403DJvNlm/c6dOnjQceeMCoXLmy4ePjY/Tt29c4dOhQgdsNJyQkGKNGjTIiIiIMNzc3IzQ01OjWrZsxZcoU55iCtjkuyOLFiy+49bBjO+U9e/YYN954oxEYGGh4enoabdu2NebMmZPvWB999JHRuXNno3LlyoaHh4dRp04d49FHHzVSUlIMw7BvI/3oo48aLVq0MPz8/AwfHx+jRYsWxvvvv3/BjCIiIhXFlVxzGIZhTJkyxYiKijK8vLwMPz8/o1mzZsZjjz1mHD161Dlm2bJlRvv27Q0vLy8jPDzceOyxx4x58+YZgLF48WLnuC5duhhNmjQ55xzDhg0zIiMjz5tt3bp1BmA8++yz5x2zf/9+AzAeeugh5zF9fHyMPXv2GD169DC8vb2NkJAQ47nnnjPy8vKcv/fdd98ZPXr0MIKDgw13d3ejRo0axt13323ExcXlO35hrlvOd6305ZdfGrVr1zbc3d2Nli1bGvPmzSvwPS9fvtyIiooy3N3dz7le27NnjzF06FAjNDTUcHNzM6pVq2Zcf/31xnfffXfez8ThQtdlXbp0cY4737+P43298cYbBR5/wYIFRseOHQ0vLy/D39/f6Nu3r7Ft27Z8Y5577jkDMI4dO3bRvCJlkcUwirgULyIiIiIiIiIichHqKSUiIiIiIiIiIiVORSkRERERERERESlxKkqJiIiIiIiIiEiJU1FKRERERERERERKnIpSIiIiIiIiIiJS4lSUEhERERERERGREudqdoCKxGazcfToUfz8/LBYLGbHERERkWJiGAanTp0iPDwcq1V/A3TQtZCIiEjFUNhrIRWlStDRo0eJiIgwO4aIiIiUkEOHDlG9enWzY5QauhYSERGpWC52LaSiVAny8/MD7P8o/v7+JqcRERGR4pKamkpERITzu1/sdC0kIiJSMRT2WkhFqRLkmKbu7++vCzEREZEKQEvU8tO1kIiISMVysWshNTkQEREREREREZESp6KUiIiIiIiIiIiUOBWlRERERERERESkxKmnlIiISAnJy8sjJyfH7BhSBNzc3HBxcTE7hoiISJmia6Hyo6iuhVSUEhERKWaGYRAfH09ycrLZUaQIBQYGEhoaqmbmIiIiF6FrofKpKK6FVJQSEREpZo6LsODgYLy9vVXEKOMMwyAjI4PExEQAwsLCTE4kIiJSuulaqHwpymshFaVERESKUV5envMirHLlymbHkSLi5eUFQGJiIsHBwVrKJyIich66FiqfiupaSI3ORUREipGjb4K3t7fJSaSoOf5N1RtDRETk/HQtVH4VxbWQilIiIiIlQNPUyx/9m4qIiBSevjfLn6L4N1VRSkRERERERERESpyKUiIiIlIiatasyYQJE8yOISIiImIaXQ/lp6KUiIiI5GOxWC54Gz9+/GUdd82aNYwcObJow4qIiIgUA10PlQztviciIiL5xMXFOe9/8803jBs3jp07dzqf8/X1dd43DIO8vDxcXS9+SVG1atWiDSoiIiJSTHQ9VDI0U0pERETyCQ0Ndd4CAgKwWCzOxzt27MDPz4/ffvuNqKgoPDw8+PPPP9mzZw/9+/cnJCQEX19f2rRpw4IFC/Id99/T1S0WC//73/+44YYb8Pb2pl69evz0008l/G5FREREzqXroZKhopSIiEgJMgyDjOxcU26GYRTZ+3jiiSd49dVX2b59O82bNyctLY3evXuzcOFC1q9fT8+ePenbty8HDx684HGef/55brrpJjZt2kTv3r259dZbSUpKKrKcIiIiUvroeii/inw9pOV7IiIiJeh0Th6Nx80z5dzbXojB271ovvpfeOEFrrvuOufjoKAgWrRo4Xz84osvMmvWLH766SdGjx593uMMHz6cIUOGAPDf//6Xd999l9WrV9OzZ88iySkiIiKlj66H8qvI10OaKSUiIiKXrHXr1vkep6Wl8cgjj9CoUSMCAwPx9fVl+/btF/3LYPPmzZ33fXx88Pf3JzExsVgyi4iIiBQlXQ9dOc2UEhERKUFebi5seyHGtHMXFR8fn3yPH3nkEWJjY3nzzTepW7cuXl5e3HjjjWRnZ1/wOG5ubvkeWywWbDZbkeUUERGR0kfXQ/lV5OshFaXKidPZeXi6WbFYLGZHERGRC7BYLEU2Zbw0WbZsGcOHD+eGG24A7H8p3L9/v7mhpELJsxnk2QzcXbUQQESktNP1kDjoW7scsNkM7pu+jpFfrCMlI8fsOCIiUgHVq1ePH374gQ0bNrBx40ZuueWWCvMXPikdZq0/Qo93lvDr5rgibWIrIiJSWLoeunQqSpUD2+JSWbb7BLHbEujz3h9sPJRsdiQREalg3n77bSpVqkSHDh3o27cvMTExXHXVVWbHkgrCMAw+X7Gf/ScyuG/6Xwz8YDmr91WMXYtERKT00PXQpbMY+lNSiUlNTSUgIICUlBT8/f2L9NibD6dw34x1HEo6jZuLhad7N2JYh5paziciYrLMzEz27dtHrVq18PT0NDuOFKEL/dsW53d+WVacn0taVi4fL93LlKV7OZ2TB8B1jUN4vGdD6gb7Fum5RESk8HQtVH4VxbWQZkqVE82qBzDn/k70bBJKTp7B+J+3cd/0v0jN1HI+ERERKf98PVx56Lr6LHm0K7e0q4GL1ULstgRiJizlqVmbSUzNNDuiiIiI/IuKUuVIgJcbH9x2Fc/1bYybi4XftsRz/bt/suVIitnRREREREpEsL8n/72hGfPGdKZH4xDybAYzVh2kx4SlHDyRYXY8EREROYuKUuWMxWLhjo61mHlPB6oFenEwKYOB7y/np41HzY4mIiIiUmLqBvsyZWhrZt4TTYMQP5Izcnjku43YbOpcISIiUlqoKFVOtYwI5NcHOnFd4xCy82w8M2szKae1lE9EREQqljY1g/h4aGu83V1YvS+JT5ftMzuSiIiInKGiVDkW4O3Gh7dFUT/El9TMXD75UxdhIiIiUvHUqOzNM30aA/D6vJ3sSjhlciIREREBFaXKPRerhTHd6wPw6Z/7OJmebXIiERERkZI3pG0EXRtUJTvXxthvN5KTZzM7koiISIWnolQF0LNJKI3C/EnLyuWjpXvNjiMiIiJS4iwWC68Nak6Alxubj6QwefFusyOJiIhUeCpKVQBWq4Wx19lnS01bvp9jp7JMTiQiIiIl6YMPPqB58+b4+/vj7+9PdHQ0v/32m/P1zMxMRo0aReXKlfH19WXQoEEkJCTkO8bBgwfp06cP3t7eBAcH8+ijj5Kbm1vSb+WKhPh78kL/JgBMWrSbTYeTzQ0kIiJSwakoVUF0bxRMi+oBnM7J48Mle8yOIyIiIiWoevXqvPrqq6xbt461a9dy7bXX0r9/f7Zu3QrAQw89xM8//8zMmTNZsmQJR48eZeDAgc7fz8vLo0+fPmRnZ7N8+XKmTZvG1KlTGTdunFlv6bL1axFOn2Zh5NoMxn67kcycPLMjiYiIVFgqSlUQFouFh87Mlvpy5QESUjNNTiQiIuVZ165dGTNmjPNxzZo1mTBhwgV/x2KxMHv27Cs+d1Edpzzp27cvvXv3pl69etSvX5+XX34ZX19fVq5cSUpKCp988glvv/021157LVFRUXz22WcsX76clStXAjB//ny2bdvGl19+ScuWLenVqxcvvvgikydPJju7bPWrtFgsvDigKVV8PdidmMab83aaHUlERMopXQ9dnIpSFUiX+lWJiqxEVq5NfRREROS8+vbtS8+ePQt87Y8//sBisbBp06ZLOuaaNWsYOXJkUcRzGj9+PC1btjzn+bi4OHr16lWk5ypP8vLy+Prrr0lPTyc6Opp169aRk5ND9+7dnWMaNmxIjRo1WLFiBQArVqygWbNmhISEOMfExMSQmprqnG1VlgT5uPPaoGYAfLJsHyv3njA5kYiIlDa6HioZKkpVIBaLhYfPzJb6evUhjiSfNjmRiIiURiNGjCA2NpbDhw+f89pnn31G69atad68+SUds2rVqnh7exdVxAsKDQ3Fw8OjRM5VlmzevBlfX188PDy45557mDVrFo0bNyY+Ph53d3cCAwPzjQ8JCSE+Ph6A+Pj4fAUpx+uO184nKyuL1NTUfLfSolujEG5qXR3DgOd+3IphGGZHEhGRUkTXQyVDRakKpkPdKrSvHUR2no1Ji3aZHUdEREqh66+/nqpVqzJ16tR8z6elpTFz5kwGDBjAkCFDqFatGt7e3jRr1oyvvvrqgsf893T1Xbt20blzZzw9PWncuDGxsbHn/M7jjz9O/fr18fb2pnbt2jz77LPk5OQAMHXqVJ5//nk2btyIxWLBYrE48/57uvrmzZu59tpr8fLyonLlyowcOZK0tDTn68OHD2fAgAG8+eabhIWFUblyZUaNGuU8V3nRoEEDNmzYwKpVq7j33nsZNmwY27ZtK9ZzvvLKKwQEBDhvERERxXq+S/V078a4u1rZmXCKrUdLT8FMRETMp+uhkrkeUlGqAnq4RwMAZq49zMETGSanERGpYAwDstPNuRVyJoirqytDhw5l6tSp+WaPzJw5k7y8PG677TaioqL45Zdf2LJlCyNHjuT2229n9erVhTq+zWZj4MCBuLu7s2rVKj788EMef/zxc8b5+fkxdepUtm3bxsSJE/n444955513ALj55pt5+OGHadKkCXFxccTFxXHzzTefc4z09HRiYmKoVKkSa9asYebMmSxYsIDRo0fnG7d48WL27NnD4sWLnU28/30RWta5u7tTt25doqKieOWVV2jRogUTJ04kNDSU7OxskpOT841PSEggNDQUsP+19d+78TkeO8YU5MknnyQlJcV5O3ToUNG+qSsU4O1G90bBAPy44YjJaUREKhBdD+l66AzXYj26lEptagbRqV4V/th1nIkLd/HWTS3MjiQiUnHkZMB/w80591NHwd2nUEPvvPNO3njjDZYsWULXrl0B+1T1QYMGERkZySOPPOIce//99zNv3jy+/fZb2rZte9FjL1iwgB07djBv3jzCw+2fxX//+99z+h4888wzzvs1a9bkkUce4euvv+axxx7Dy8sLX19fXF1dL1gUmTFjBpmZmXz++ef4+Njf+6RJk+jbty+vvfaacwlapUqVmDRpEi4uLjRs2JA+ffqwcOFC7rrrrkJ9XmWRzWYjKyuLqKgo3NzcWLhwIYMGDQJg586dHDx4kOjoaACio6N5+eWXSUxMJDjYXsSJjY3F39+fxo0bn/ccHh4epX7pQP+W1fh1czw/bTzKE70a4WK1mB1JRKT80/WQrofO0EypCsoxW2rW+sPsOZZ2kdEiIlLRNGzYkA4dOvDpp58CsHv3bv744w9GjBhBXl4eL774Is2aNSMoKAhfX1/mzZvHwYMHC3Xs7du3ExER4bwAA5zFj7N98803dOzYkdDQUHx9fXnmmWcKfY6zz9WiRQvnBRhAx44dsdls7Nz5z65rTZo0wcXFxfk4LCyMxMTESzpXafbkk0+ydOlS9u/fz+bNm3nyySf5/fffufXWWwkICGDEiBGMHTuWxYsXs27dOu644w6io6Np3749AD169KBx48bcfvvtbNy4kXnz5vHMM88watSoUl90upiuDaoS4OVGQmqWGp6LiEg+uh4q/ushzZSqoFpGBHJtw2AW7Ujku3WHebxnQ7MjiYhUDG7e9r/QmXXuSzBixAjuv/9+Jk+ezGeffUadOnXo0qULr732GhMnTmTChAk0a9YMHx8fxowZQ3Z2dpFFXbFiBbfeeivPP/88MTExBAQE8PXXX/PWW28V2TnO5ubmlu+xxWLBZrMVy7nMkJiYyNChQ4mLiyMgIIDmzZszb948rrvuOgDeeecdrFYrgwYNIisri5iYGN5//33n77u4uDBnzhzuvfdeoqOj8fHxYdiwYbzwwgtmvaUi4+HqQu9mYXy1+iCz1x+hY90qZkcSESn/dD1UKBXhekhFqQqsd7MwFu1I1F8FRURKksVS6CnjZrvpppt48MEHmTFjBp9//jn33nsvFouFZcuW0b9/f2677TbAvgzs77//vuAyrrM1atSIQ4cOERcXR1hYGAArV67MN2b58uVERkby9NNPO587cOBAvjHu7u7k5eVd9FxTp04lPT3d+dfBZcuWYbVaadCgQaHylgeffPLJBV/39PRk8uTJTJ48+bxjIiMj+fXXX4s6WqkwoGU4X60+yNwt8bw4oCmebi4X/yUREbl8uh7S9dAZWr5XgbWvHQTApsMppGXlmpxGRERKG19fX26++WaefPJJ4uLiGD58OAD16tUjNjaW5cuXs337du6+++5zmmBfSPfu3alfvz7Dhg1j48aN/PHHH/kuthznOHjwIF9//TV79uzh3XffZdasWfnG1KxZk3379rFhwwaOHz9OVlbWOee69dZb8fT0ZNiwYWzZsoXFixdz//33c/vttzv7J4i0qRlEtUAvTmXlsnB7+Vm2KSIiV07XQ8VLRakKrHolbyKCvMizGazdn2R2HBERKYVGjBjByZMniYmJcfY8eOaZZ7jqqquIiYmha9euhIaGMmDAgEIf02q1MmvWLE6fPk3btm35v//7P15++eV8Y/r168dDDz3E6NGjadmyJcuXL+fZZ5/NN2bQoEH07NmTa665hqpVqxa4DbO3tzfz5s0jKSmJNm3acOONN9KtWzcmTZp06R+GlFtWq4V+Le3/+56tXfhERORfdD1UfCyGUcj9EOWKpaamEhAQQEpKCv7+/mbHAeDRmRuZue4wd3epzZO9GpkdR0Sk3MnMzGTfvn3UqlULT09Ps+NIEbrQv21p/M4vDUrz57Iz/hQxE5bi5mJhzdPdCfR2NzuSiEi5oGuh8qsoroU0U6qCi65TGYCVezVTSkRERCquBqF+NArzJyfP4JfNcWbHERERqRBUlKrg2te2F6W2HEnhVGaOyWlEREREzDPgzBK+H9ebtCOUiIhIBaOiVAUXHuhFZGXvM32lTpodR0RERMQ0/VqGY7HA6v1JHD6ZYXYcERGRck9FKaF9LftsqRV7T5icRERERMQ8YQFezuuiHzdotpSIiEhxU1FKzuorpaKUiIiIVGwDWp3ZhW/9EbQfkIiISPFSUUry9ZVKVV8pEZFiYbPZzI4gRUz/puVTz6ZhuLta2ZWYxra4VLPjiIiUG/reLH+K4t/UtQhySBkXGuBJrSo+7Duezpp9SXRrFGJ2JBGRcsPd3R2r1crRo0epWrUq7u7uWCwWs2PJFTAMg+zsbI4dO4bVasXd3d3sSFKEArzc6NYwmN+2xPPjhqM0CQ8wO5KISJmma6HypyivhVSUEgDa1w5i3/F0Vuw5oaKUiEgRslqt1KpVi7i4OI4eVY+a8sTb25saNWpgtWrieXkzoFW1M0WpIzzesyEuVv2fJxGRy6VrofKrKK6FVJQSwL6E76vVh1i5T32lRESKmru7OzVq1CA3N5e8vDyz40gRcHFxwdXVVX/pLae6NqiKv6crCalZrNp7gg51q5gdSUSkTNO1UPlTVNdCKkoJANFn+kptPZpKSkYOAd5uJicSESlfLBYLbm5uuLnpv68ipZ2Hqwt9mofx1epDzN0ar6KUiEgR0LWQFETzzQWAYH9Palf1wTBg9f4ks+OIiIiImKpzvaoArN6n6yIREZHioqKUODl24VuxR0v4REREpGJrXTMIgB3xp0jOyDY5jYiISPmkopQ4OZbwrdyropSIiIhUbFX9PKhd1QeANftPmpxGRESkfFJRSpza1bb/RXB7fKr+IigiIiIVXrta9mujNWptICIiUixUlBKnYD9P6gb7YhiwSv0TREREpIJrc2YJn66LREREioeKUpJP+zOzpbSET0RERCq6tmdmSm09kkJ6Vq7JaURERMofFaUkn+ja9i2P1excREREKrrqlbwJD/Ak12aw/mCy2XFERETKHRWlJB9HX6kd8ac4ma6+UiIiIlKxOWZLrVZfKRERkSKnopTkU8XXg/ohvgCs2qfZUiIiIlKxtXEUpXRdJCIiUuRUlJJztK9dGYCVe/UXQREREanYHDvwrT+YTHauzeQ0IiIi5YuKUnKO6DNFKfWVEhERkYquTlVfgnzcycq1sflIstlxREREyhUVpeQc7c4UpXYmnOJEWpbJaURERETMY7FYaFOzEgCr9500OY2IiEj5YmpR6pVXXqFNmzb4+fkRHBzMgAED2LlzZ74xmZmZjBo1isqVK+Pr68ugQYNISEjIN+bgwYP06dMHb29vgoODefTRR8nNzb9t7++//85VV12Fh4cHdevWZerUqefkmTx5MjVr1sTT05N27dqxevXqS85SHgT5uFOrig9gb3guIiIiUpG1rWX/g536SomIiBQtU4tSS5YsYdSoUaxcuZLY2FhycnLo0aMH6enpzjEPPfQQP//8MzNnzmTJkiUcPXqUgQMHOl/Py8ujT58+ZGdns3z5cqZNm8bUqVMZN26cc8y+ffvo06cP11xzDRs2bGDMmDH83//9H/PmzXOO+eabbxg7dizPPfccf/31Fy1atCAmJobExMRCZylP6lS1F6X2HkszOYmIiIiIudrWtPeVWnvgJHk2w+Q0IiIi5YfFMIxS88167NgxgoODWbJkCZ07dyYlJYWqVasyY8YMbrzxRgB27NhBo0aNWLFiBe3bt+e3337j+uuv5+jRo4SEhADw4Ycf8vjjj3Ps2DHc3d15/PHH+eWXX9iyZYvzXIMHDyY5OZm5c+cC0K5dO9q0acOkSZMAsNlsREREcP/99/PEE08UKsvFpKamEhAQQEpKCv7+/kX62RW1V37dzkdL9zK8Q03G92tidhwREZEypSx955eksvq55ObZaPlCLGlZufzywNU0CQ8wO5KIiEipVtjv/FLVUyolJQWAoCD7X6PWrVtHTk4O3bt3d45p2LAhNWrUYMWKFQCsWLGCZs2aOQtSADExMaSmprJ161bnmLOP4RjjOEZ2djbr1q3LN8ZqtdK9e3fnmMJkKU9qn5kptUczpURERKSCc3WxclWko6+UdicWEREpKqWmKGWz2RgzZgwdO3akadOmAMTHx+Pu7k5gYGC+sSEhIcTHxzvHnF2QcrzueO1CY1JTUzl9+jTHjx8nLy+vwDFnH+NiWf4tKyuL1NTUfLeyok5VXwD2Hku/yEgRERGR8q9dLfsfTdfsV1FKRESkqJSaotSoUaPYsmULX3/9tdlRiswrr7xCQECA8xYREWF2pEKrfaYodST5NKez80xOIyIiImKutmeKUqv3JVGKul+IiIiUaaWiKDV69GjmzJnD4sWLqV69uvP50NBQsrOzSU5Ozjc+ISGB0NBQ55h/74DneHyxMf7+/nh5eVGlShVcXFwKHHP2MS6W5d+efPJJUlJSnLdDhw4V4tMoHYJ83An0dgNg33HNlhIREZGKrXn1ANxdrRxPy9a1kYiISBExtShlGAajR49m1qxZLFq0iFq1auV7PSoqCjc3NxYuXOh8bufOnRw8eJDo6GgAoqOj2bx5c75d8mJjY/H396dx48bOMWcfwzHGcQx3d3eioqLyjbHZbCxcuNA5pjBZ/s3DwwN/f/98t7LEsYRPfaVERESkovNwdaFlRCCgvlIiIiJFxdSi1KhRo/jyyy+ZMWMGfn5+xMfHEx8fz+nTpwEICAhgxIgRjB07lsWLF7Nu3TruuOMOoqOjnbvd9ejRg8aNG3P77bezceNG5s2bxzPPPMOoUaPw8PAA4J577mHv3r089thj7Nixg/fff59vv/2Whx56yJll7NixfPzxx0ybNo3t27dz7733kp6ezh133FHoLOVN7Sr2ZufqKyUiIiICbWueWcKnvlIiIiJFwtXMk3/wwQcAdO3aNd/zn332GcOHDwfgnXfewWq1MmjQILKysoiJieH99993jnVxcWHOnDnce++9REdH4+Pjw7Bhw3jhhRecY2rVqsUvv/zCQw89xMSJE6levTr/+9//iImJcY65+eabOXbsGOPGjSM+Pp6WLVsyd+7cfM3PL5alvKkTrJlSIiIiIg5tawXBYs2UEhERKSoWQ50aS0xqaioBAQGkpKSUiaV887fGM/KLdTSt5s+c+zuZHUdERKTMKGvf+SWlrH8uaVm5NB8/D5sBy5+4lvBAL7MjiYiIlEqF/c4vFY3OpXRyzJTaeyxdu8yIiIhIhefr4UrTagEArNESPhERkSumopScV40gb1ytFjKy84hPzTQ7joiIiIjpnH2ltIRPRETkiqkoJefl5mKlRpA3oGbnIiIiIgBtaqkoJSIiUlRUlJILql1Vzc5FREREHNqcmSm1KzGNpPRsk9OIiIiUbSpKyQXVqeoDaKaUiIiICECQjzv1zvTdXHfgpMlpREREyjYVpeSC6mimlIiIiEg+LSMCAdh0ONnUHCIiImWdilJyQbU1U0pEREQknxZnilIbDiWbmkNERKSsU1FKLsgxU+pI8mlOZ+eZnEZERETEfI6ZUhsPJWMYhrlhREREyjAVpeSCKvm4U8nbDYC9x7WET0RERKRBqB8erlZSM3PZd1yzyUVERC6XilJyUY4d+LSET0RERATcXKw0rRYAwEb1lRIREblsKkrJRTl24FOzcxERERG7FtUDAdh4KMXcICIiImWYilJyUZopJSIiIpJfiwj7TCk1OxcREbl8KkrJRTmanaunlIiIiIido9n5tqOpZOfazA0jIiJSRqkoJRdV+8zyvb3H0rXDjIiIiAhQI8ibQG83svNs7IhPNTuOiIhImaSilFxUjSBvXK0WMrLziE/NNDuOiIiIiOksFstZfaWSTc0iIiJSVqkoJRfl5mKlRmVvAPYkqq+UiIiICECLM0v4NqjZuYiIyGVRUUoKpXYV9ZUSEREROVvLM83ONx5ONjeIiIhIGaWilBRKnWB7X6k9iSpKiYiIiADO5Xt7jqWRmpljbhgREZEySEUpKZQ6zplSWr4nIiIiAlDZ14OIIC8MA7Yc1hI+ERGRS6WilBSKY6bU3mMqSomIiIg4OGZLrVezcxERkUumopQUiqOn1JHk02Rk55qcRkRERKR0aHmm2bl24BMREbl0KkpJoVTycSfIxx2AfVrCJyIiIgL8swOfmp2LiIhcOhWlpNBqVznT7FxL+EREREQAaBLuj4vVQkJqFvEpmWbHERERKVNUlJJCq13V0VdKO/CJiIiIAHi7u1I/xA+ADVrCJyIicklUlJJCq1PV3ldKM6VERERE/tEyIgDQEj4REZFLpaKUFFrtM0UpzZQSERER+YdjBz41OxcREbk0KkpJodVxLt9Lx2YzTE4jIiIiUjq0rBEIwKbDKbpGEhERuQQqSkmhRQR542q1cDonj/hUNfIUERERAagX7Ie3uwtpWbnsPa4Z5SIiIoWlopQUmpuLlcjK3oB9tpSIiIiIgIvVQtNq9r5SGw6lmJxGRESk7FBRSi5JbWezc/0VUERERMShZUQgoL5SIiIil0JFKbkktZ19pVSUEhEREXFwNjvXDnwiIiKFpqKUXJI6zplSWr4nIiIi4tAiwr58b3tcKpk5eSanERERKRtUlJJL4tiBb99xFaVEREREHKoFelHF152cPINtcalmxxERESkTVJSSS1K9kr3ReXxqJrl5NpPTiIiIiJQOFovlnyV86islIiJSKCpKySWp6uuBm4uFPJtB4qkss+OIiIiIlBot1OxcRETkkqgoJZfEarUQFuAFwNHk0yanERERESk9HEWpDSpKiYiIFIqKUnLJwgM9ATiiopSIiIiIU8szy/f2n8jgRJpmlIuIiFyMilJyycIDHTOlMk1OIiIiIoXxyiuv0KZNG/z8/AgODmbAgAHs3Lkz35iuXbtisVjy3e655558Yw4ePEifPn3w9vYmODiYRx99lNzc3JJ8K6VagLcb9YLtOxWvO3DS5DQiIiKln4pScsmqnSlKHUnOMDmJiIiIFMaSJUsYNWoUK1euJDY2lpycHHr06EF6ev7ddO+66y7i4uKct9dff935Wl5eHn369CE7O5vly5czbdo0pk6dyrhx40r67ZRqrWtWAlSUEhERKQxXswNI2aOZUiIiImXL3Llz8z2eOnUqwcHBrFu3js6dOzuf9/b2JjQ0tMBjzJ8/n23btrFgwQJCQkJo2bIlL774Io8//jjjx4/H3d29WN9DWREVGcRXqw+xVkUpERGRi9JMKblk/xSl1FNKRESkLEpJSQEgKCgo3/PTp0+nSpUqNG3alCeffJKMjH9mRa9YsYJmzZoREhLifC4mJobU1FS2bt1a4HmysrJITU3NdyvvWkfaZ0ptPpxCZk6eyWlERERKN82UkktWTY3ORUREyiybzcaYMWPo2LEjTZs2dT5/yy23EBkZSXh4OJs2beLxxx9n586d/PDDDwDEx8fnK0gBzsfx8fEFnuuVV17h+eefL6Z3UjpFVvamiq87x9Oy2XIkhdY1gy7+SyIiIhWUilJyyRwzpU5l5pKamYO/p5vJiURERKSwRo0axZYtW/jzzz/zPT9y5Ejn/WbNmhEWFka3bt3Ys2cPderUuaxzPfnkk4wdO9b5ODU1lYiIiMsLXkZYLBaiIisxb2sC6w6cVFFKRETkArR8Ty6Zt7srlbzthag49ZUSEREpM0aPHs2cOXNYvHgx1atXv+DYdu3aAbB7924AQkNDSUhIyDfG8fh8fag8PDzw9/fPd6sIWkfaC1HqKyUiInJhKkrJZVFfKRERkbLDMAxGjx7NrFmzWLRoEbVq1bro72zYsAGAsLAwAKKjo9m8eTOJiYnOMbGxsfj7+9O4ceNiyV1WRZ3Zge+vAycxDMPkNCIiIqWXilJyWRxFqcMqSomIiJR6o0aN4ssvv2TGjBn4+fkRHx9PfHw8p0/bv8f37NnDiy++yLp169i/fz8//fQTQ4cOpXPnzjRv3hyAHj160LhxY26//XY2btzIvHnzeOaZZxg1ahQeHh5mvr1Sp2l4AB6uVk6kZ7PveLrZcUREREotFaXkslTTTCkREZEy44MPPiAlJYWuXbsSFhbmvH3zzTcAuLu7s2DBAnr06EHDhg15+OGHGTRoED///LPzGC4uLsyZMwcXFxeio6O57bbbGDp0KC+88IJZb6vUcne10qJ6IKAlfCIiIheiRudyWcLP7MCnopSIiEjpd7ElZBERESxZsuSix4mMjOTXX38tqljlWlTNSqzen8S6/Se5qXX5bu4uIiJyuTRTSi6LekqJiIiInF9UDXtfqbUHkkxOIiIiUnqpKCWX5Z+ilHbfExEREfm3qEh7UWrPsXROpmebnEZERKR0UlFKLkv1M0Wp+NRMcvNsJqcRERERKV0q+bhTp6oPAOvUV0pERKRAKkrJZani64Gbi4U8m0HCqSyz44iIiIiUOq0jgwA1OxcRETkfFaXkslitFsIC1FdKRERE5HyiatqX8K1TXykREZECqSgll0078ImIiIicX+szfaU2Hk4hKzfP5DQiIiKlj4pSctkczc6PqCglIiIico5aVXyo7ONOdq6NLUdSzY4jIiJS6qgoJZetWqCW74mIiIicj8Vi4apILeETERE5HxWl5LKFO4tSmSYnERERESmdHEv41u5Xs3MREZF/U1FKLptmSomIiIhcWGtns/OTGIZhchoREZHSRUUpuWzOnlInVZQSERERKUjTagG4u1o5kZ7N/hMZZscREREpVVSUksvm2H3vVFYuqZk5JqcRERERKX08XF1oXi0AgLX71VdKRETkbCpKyWXzdnelkrcboCV8IiIiIucTddYSPhEREfmHilJyRcLVV0pERETkglpHBgGwVkUpERGRfFSUkivi7CulHfhEREREChR1Zge+3YlpJGdkm5xGRESk9FBRSq6IduATERERubAgH3dqV/UBtIRPRETkbCpKyRVxNDtXUUpERETk/FqfmS21Zr+KUiIiIg4qSskVqRboDcCRkypKiYiIiJxPu1qVAVi8I9HkJCIiIqWHilJyRTRTSkREROTiujcKwc3Fws6EU+xKOGV2HBERkVJBRSm5Io6eUvGpmeTm2UxOIyIiIlI6BXi7cXXdKgDM2RRnchoREZHSQUUpuSJVfD1wc7FgMyDhVJbZcURERERKreubhwPwy+Y4DMMwOY2IiIj5VJSSK2K1WggL0A58IiIiIhdzXZMQ3F2s7E5MY6eW8ImIiKgoJVdOfaVERERELs7f043O9asC8IuW8ImIiKgoJVcu/ExfqSMqSomIiIhcUN8WYYC9r5SW8ImISEWnopRcseqOotRJFaVERERELqRboxA8XK3sO57O1qOpZscRERExlYpScsUcM6W0fE9ERETkwnw9XLmmQTBgb3guIiJSkakoJVfsn6JUpslJREREREq/651L+I5qCZ+IiFRoKkrJFdNMKREREZHCu7ZhMF5uLhxKOs3mIylmxxERETGNilJyxRy7753KyiU1M8fkNCIiIiKlm7e7K9c2si/hm6Nd+EREpAJTUUqumLe7K5W83QDNlhIREREpjL7N7Uv4ftEufCIiUoGpKCVFQkv4RERERAqva4NgfNxdOJJ8mvWHks2OIyIiYgoVpaRIOIpSR06qKCUiIiJyMZ5uLnRvHALAnI1awiciIhWTilJSJKo5ilLagU9ERESkUK5vHg7Ar5vjsNm0hE9ERCoeFaWkSFTT8j0RERGRS9K5fhX8PFyJT81k3cGTZscREREpcSpKSZFQTykRERGRS+Ph6sJ1TexL+H7RLnwiIlIBqSglRSI80BNQUUpERETkUlzv2IVvcxx5WsInIiIVjIpSUiQcy/fiUzPJzbOZnEZERESkbLi6blUCvNw4diqLZ2ZvJj5F/TlFRKTiUFFKikQVXw/cXCzYDEg4lWV2HBEREZEywd3Vyp0dawHw1epDdH59Mc/9uEXFKRERqRBUlJIiYbVaCAs4swPfSS3hExERESmsB7rVZcZd7WhbM4jsPBvTVhxQcUpERCoEV7MDSPlRLdCLg0kZ6islIiIicgksFgsd6lQhunZlVuw9wYTYXazen8S0FQf4avUhrm0YTIi/B0E+HgT5ulPZx50gH3eq+LpTu4ovVqvF7LcgIiJyWVSUkiLj2IHviIpSIiIiIpcsX3FqzwkmLLAXp+ZujT/v7zQM9eONG1vQrHpACSYVEREpGipKSZGpph34RERERK6YxWKhQ90qRNepzJr9J9l0OJmTGdkkpWdzIs3+Myk9m6Mpp9kRf4oB7y/jrk61GdO9Hp5uLmbHFxERKTQVpaTIhJ7pKaXeByIiIiJXzmKx0LZWEG1rBRX4+om0LMb/vI2fNx7lwyV7mL81ntdubE6bmkFgGGDRsj4RESnd1OhcikxYgH2mVJyKUiIiIiLFrrKvB+8NacXHQ1sT7OfB3uPp3PLRHyyffDe2N+vDzt/MjigiInJBKkpJkQk9U5SKT1VRSkRERKSkXNc4hNixXbijpR/TXF+lw7GvsaYnkvfDPXDq/P2oREREzKailBSZ8DPL95LSs8nMyTM5jYiIiEjFEZCyk+fiR9HBZRsZeLLXFopLVjJZP9xnX8onIiJSCqkoJUXG38sVrzPNNRM0W0pERESkZGydBZ9cB8kHoVIt8u5cwEs+T5NluOGxbyE5qz81O6GIiEiBTC1KLV26lL59+xIeHo7FYmH27Nn5Xh8+fDgWiyXfrWfPnvnGJCUlceutt+Lv709gYCAjRowgLS0t35hNmzbRqVMnPD09iYiI4PXXXz8ny8yZM2nYsCGenp40a9aMX3/9Nd/rhmEwbtw4wsLC8PLyonv37uzatatoPohywmKxOPtKHU1WUUpERESkWNlssOglmDkccjKg9jVw1yL8ajTjmTsH8q5liH3Y3KcwTuwxN6uIiEgBTC1Kpaen06JFCyZPnnzeMT179iQuLs55++qrr/K9fuutt7J161ZiY2OZM2cOS5cuZeTIkc7XU1NT6dGjB5GRkaxbt4433niD8ePHM2XKFOeY5cuXM2TIEEaMGMH69esZMGAAAwYMYMuWLc4xr7/+Ou+++y4ffvghq1atwsfHh5iYGDIzVXw52z99pU6bnERERESkHMvNhm9uhaVv2B9Hj4ZbvwNv+059tav60v6WZ1hpa4yHkUnCtOFgU3sFEREpXVzNPHmvXr3o1avXBcd4eHgQGhpa4Gvbt29n7ty5rFmzhtatWwPw3nvv0bt3b958803Cw8OZPn062dnZfPrpp7i7u9OkSRM2bNjA22+/7SxeTZw4kZ49e/Loo48C8OKLLxIbG8ukSZP48MMPMQyDCRMm8Mwzz9C/f38APv/8c0JCQpg9ezaDBw8uqo+kzAvVDnwiIiIixW/zTNj5K7h4QL/3oMXN5wzpVD+E77q8xamlAwlN3cTuWS9Rd9BzJoQVEREpWKnvKfX7778THBxMgwYNuPfeezlx4oTztRUrVhAYGOgsSAF0794dq9XKqlWrnGM6d+6Mu7u7c0xMTAw7d+7k5MmTzjHdu3fPd96YmBhWrFgBwL59+4iPj883JiAggHbt2jnHFCQrK4vU1NR8t/LOsXwvXkUpERERkeKzY479Z6eHCyxIOQy6Npp5NR4CIHLTRPZvWVkS6URERAqlVBelevbsyeeff87ChQt57bXXWLJkCb169SIvzz71OD4+nuDg4Hy/4+rqSlBQEPHx8c4xISEh+cY4Hl9szNmvn/17BY0pyCuvvEJAQIDzFhERcUnvvywKPbMDn2ZKiYiIiBST7HTYs8h+v2GfCw61WCz0H/YIazw74GbJI+/7uziRnFICIUVERC6uVBelBg8eTL9+/WjWrBkDBgxgzpw5rFmzht9//93saIXy5JNPkpKS4rwdOnTI7EjFLsxfM6VEREREitWexZCbCYGRENLkosPdXF2oN+ITkgigjnGQZVMeIjfPVgJBRURELqxUF6X+rXbt2lSpUoXdu3cDEBoaSmJiYr4xubm5JCUlOftQhYaGkpCQkG+M4/HFxpz9+tm/V9CYgnh4eODv75/vVt6FBaqnlIiIiEix2vGL/WfDPmCxFOpXAquGk9nrHQCuT/+BeXN/LK50IiIihVamilKHDx/mxIkThIWFARAdHU1ycjLr1q1zjlm0aBE2m4127do5xyxdupScnBznmNjYWBo0aEClSpWcYxYuXJjvXLGxsURHRwNQq1YtQkND841JTU1l1apVzjFiF3Zm+d7xtCyycrXDi4iIiEiRysuFv+fa719k6d6/hbcbxN5q/bBaDOqvfpZjyaeKIaCIiEjhmVqUSktLY8OGDWzYsAGwNxTfsGEDBw8eJC0tjUcffZSVK1eyf/9+Fi5cSP/+/albty4xMTEANGrUiJ49e3LXXXexevVqli1bxujRoxk8eDDh4eEA3HLLLbi7uzNixAi2bt3KN998w8SJExk7dqwzx4MPPsjcuXN566232LFjB+PHj2ft2rWMHj0asK/FHzNmDC+99BI//fQTmzdvZujQoYSHhzNgwIAS/cxKu0rebri72v9nlZiaZXIaERERkXLm0Eo4nQReQRDR/pJ/PXLIBFIs/tSzHGLtjPFFn09EROQSmFqUWrt2La1ataJVq1YAjB07llatWjFu3DhcXFzYtGkT/fr1o379+owYMYKoqCj++OMPPDw8nMeYPn06DRs2pFu3bvTu3Zurr76aKVOmOF8PCAhg/vz57Nu3j6ioKB5++GHGjRvHyJEjnWM6dOjAjBkzmDJlCi1atOC7775j9uzZNG3a1Dnmscce4/7772fkyJG0adOGtLQ05s6di6enZwl8UmWHxWJx7sCnJXwiIiIiRcyxdK9+T3BxveRfd/GtzMnOLwBwbcI0tmxad5HfEBERKT4WwzAMs0NUFKmpqQQEBJCSklKu+0vd/NEKVu1LYuLglvRvWc3sOCIiIiWuonznXyp9LlfIMGBiC0g+ADdPh0bXX/Zxdr7VgwZpq9nk2owmTy7FxaVMdfUQEZFSrrDf+fr2kSLnmCmlHfhEREREilDCVntBytUT6lxz+cexWKg6eDKncad57mZWzXqv6DKKiIhcAhWlpMiFnml2ruV7IiIiIkVo56/2n3WuBXefKzpUUPX6bKk3CoAmW17jZOKRK00nIiJyyVSUkiIXHqiZUiIiIiJFbscc+89L3HXvfFrd9BS7XWoTQDoHZjxYJMcUERG5FCpKSZEL9Xc0Oj9tchIRERGRciL5EMRtBIvV3uS8CLi6uZPV8x3yDAstk2PZu2J2kRxXRESksFSUkiIXpuV7IiIiIkVr52/2nxHtwadKkR22SZuu/BF0IwA+sY9hy0wrsmOLiIhcjIpSUuRCzzQ6P5aWRU6ezeQ0IiIiIuWAc+le7yI/dONbX+OIUZUQWwK7vnmqyI8vIiJyPipKSZGr7OOOm4sFw4DEU1lmxxEREREp206fhP1/2u83KPqiVHCVymxqOc5++H3TSN/yW5GfQ0REpCAqSkmRs1othPg7mp2rr5SIiIjIFdkVC0YeBDeGynWK5RTd+t7Gj269ALDMuhtSDhfLeURERM6mopQUi7AAR7Nz9ZUSERERuSKOpXvFMEvKwd3VStDAN9lkq4V3XgqnZwyFvJxiO5+IiAioKCXFxNHsPF5FKREREZHLl5MJuxfa7zfsU6yn6tSoOl9HvkCq4Y1XwjqMBc8X6/lERERUlJJi4ZgpdTRZRSkRERGRy7ZvKWSngV84hLcq9tPdM6A7T9ruAcCy4j3Y8Uuxn1NERCouFaWkWDh24ItPVU8pERERkct29q57Fkuxn65GZW9qdxrM/3Lt/aWM2ffCyf3Ffl4REamYVJSSYqGeUiIiIiJXyDDg77n2+8W8dO9s93Wty+few1lvq4slMwVm3gG52lFZRESKnopSUixC1VNKRERE5MqkJUJaAlisENmxxE7r5e7CY9c3Z1T2A5w0fOHoXzD/2RI7v4iIVBwqSkmxcMyUSjyVRW6ezeQ0IiIiFdsrr7xCmzZt8PPzIzg4mAEDBrBz5858YzIzMxk1ahSVK1fG19eXQYMGkZCQkG/MwYMH6dOnD97e3gQHB/Poo4+Sm5tbkm+lYkk+YP/pFw6uHiV66j7NwqhRuz5jc+61P7H6I/WXEhGRIqeilBSLKr4euFgt5NkMjqdlmx1HRESkQluyZAmjRo1i5cqVxMbGkpOTQ48ePUhPT3eOeeihh/j555+ZOXMmS5Ys4ejRowwcOND5el5eHn369CE7O5vly5czbdo0pk6dyrhx48x4SxXDyTNFqUqRJX5qi8XC+H5NWGK04uPc3vYn/3irxHOIiEj5pqKUFAsXq4UQP/tf9OJS1OxcRETETHPnzmX48OE0adKEFi1aMHXqVA4ePMi6desASElJ4ZNPPuHtt9/m2muvJSoqis8++4zly5ezcuVKAObPn8+2bdv48ssvadmyJb169eLFF19k8uTJZGfrD1DFwjFTKrDki1IADUP9ub19JB/m9iUHVziyDuI2mZJFRETKJxWlpNiEBdr7SqnZuYiISOmSkpICQFBQEADr1q0jJyeH7t27O8c0bNiQGjVqsGLFCgBWrFhBs2bNCAkJcY6JiYkhNTWVrVu3lmD6CiTZvJlSDg9dVx+bdxXm5bW2P7FuqmlZRESk/FFRSopNqHbgExERKXVsNhtjxoyhY8eONG3aFID4+Hjc3d0JDAzMNzYkJIT4+HjnmLMLUo7XHa8VJCsri9TU1Hw3uQSO5XuBNUyLEOjtziMxDZie1w0AY9M3kJVmWh4RESlfVJSSYhPmby9KxWv5noiISKkxatQotmzZwtdff13s53rllVcICAhw3iIiIor9nOWKycv3HG5uHcGJKm3ZawvFkp0GW743NY+IiJQfKkpJsdFMKRERkdJl9OjRzJkzh8WLF1O9enXn86GhoWRnZ5OcnJxvfEJCAqGhoc4x/96Nz/HYMebfnnzySVJSUpy3Q4cOFeG7KedseZBy2H7fxOV7AK4uVp7q05iv8q4FIGvVJ6bmERGR8kNFKSk2YQH2nlLxKkqJiIiYyjAMRo8ezaxZs1i0aBG1atXK93pUVBRubm4sXLjQ+dzOnTs5ePAg0dHRAERHR7N582YSExOdY2JjY/H396dx48YFntfDwwN/f/98Nymk1KNgywWrG/iFmZ2Grg2CORI5gCzDFY/EjXB0g9mRRESkHFBRSoqNZkqJiIiUDqNGjeLLL79kxowZ+Pn5ER8fT3x8PKdP25fYBwQEMGLECMaOHcvixYtZt24dd9xxB9HR0bRv3x6AHj160LhxY26//XY2btzIvHnzeOaZZxg1ahQeHh5mvr3yybl0LwKsLuZmOeOBftHMt7UBIPH3D01OIyIi5YGKUlJsws4UpRJSM8mzGSanERERqbg++OADUlJS6Nq1K2FhYc7bN9984xzzzjvvcP311zNo0CA6d+5MaGgoP/zwg/N1FxcX5syZg4uLC9HR0dx2220MHTqUF154wYy3VP6Vgibn/9Yw1J+4ekMA8Ns1G9tpNa4XEZEr42p2ACm/gv08sFog12ZwIi2L4DONz0VERKRkGcbF/zjk6enJ5MmTmTx58nnHREZG8uuvvxZlNDmf5IP2nyY3Of+3AQNuYu9br1Kbo2yY+z9a3jDW7EgiIlKGaaaUFBtXFyvBflrCJyIiInLJHMv3TG5y/m/B/l4cqX0TAJ6bviAzJ8/kRCIiUpapKCXFSn2lRERERC6Dc/le6SpKAbTuP5psXGlo7GXOb5o5JyIil09FKSlWjr5S8SmnTU4iIiIiUoYkl96ilFdgVeKrxdgfrPuUE2lZ5gYSEZEyS0UpKVbOmVKpmiklIiIiUii52ZB61H6/lC3fc6je/T4AerGMD+ZtMDeMiIiUWSpKSbH6Z6aUilIiIiIihZJyCDDAzRt8qpqdpkDWmh3JCKiLjyWLrPVfszsxzexIIiJSBqkoJcUqNMALgLhkFaVERERECsW5dK8GWCzmZjkfiwXv9iMAGGJdyDvzd5ocSEREyiIVpaRYhTmX76mnlIiIiEihnDyrKFWatRiMzcWDxtYDxG9dwpYjKWYnEhGRMkZFKSlWjqJUQkoWNpthchoRERGRMiD5oP1nKWxyno93ENbmNwHwoOsPvDFPs6VEROTSqCglxSrYzxOLBbLzbCRlZJsdR0RERKT0cyzfK6VNzvPp/AiG1ZXOLptJ2/Unq/aeMDuRiIiUISpKSbFyd7VSxdcDULNzERERkUJxLt8rA0WpSjWxtLwFgIdcv+PN+TsxDM2OFxGRwlFRSoqds6+UilIiIiIiF5dcRnpKOXR6BMPqxtUuW7EcWM7vfx8zO5GIiJQRKkpJsQv1txel4lPU7FxERETkgrIzIP1MUacsLN8DqBSJpdVtADzk+j1vzN2pXqIiIlIoKkpJsXPMlDqqmVIiIiIiF+Zocu4RAF6VzM1yKTo9jGF1I9plG/4JK/l1S5zZiUREpAxQUUqKXWiAF6CeUiIiIiIX5WxyXkaW7jkERmCJGgbYe0u9PW8nuXk2k0OJiEhpp6KUFLvwQEdPKS3fExEREbmgstTk/N+uHovh4k476w5CTq7hh7+OmJ1IRERKORWlpNj901NKM6VERERELii5DBelAqphiRoOwFjXmUyI3UlmTp65mUREpFRTUUqKXdiZ5XtxKZnaIlhERETkQpzL98pgUQrOzJbyoI31b2qnrWXGqoNmJxIRkVJMRSkpdsH+HgBk5dpIzsgxOY2IiIhIKVaWl+8B+IdhaX0nYO8t9f7iXZotJSIi56WilBQ7TzcXKvu4A/bZUiIiIiJyHmV9phTA1WMwXD2Jsu6i8em1/LhBvaVERKRgKkpJiQg70+z8SLKanYuIiIgU6HQyZKbY7wdEmBrliviFYmk9AoAxrt/zyZ/71MJBREQKpKKUlIiISt4AHEzKMDmJiIiISCmVfKb/kncV8PA1N8uVunoMhtWNq6y7yUvcyR+7jpudSERESiEVpaRE1Kh8pih1It3kJCIiIiKlVHlYuufgG4ylzjUA9Lau4pM/95kcSERESiMVpaRERAb5AHBAM6VEREREClbWm5z/W+MBAPRxWcWSv4+xO/GUuXlERKTUUVFKSkRkZS3fExEREbkgx0ypwBrm5igqDXuD1Y2G1kPUsRzhkz/3m51IRERKGRWlpETUCLIXpQ4nnSbPpkaXIiIiIudw9JQqD8v3ALwqwZklfH2sq/jhr8MkpWebHEpEREoTFaWkRIQHeuFqtZCdZyM+NdPsOCIiIiKlT3lbvgfQ5AYABnquISvXxvSVB0wOJCIipYmKUlIiXKwWqlfyAuCAmp2LiIiI5GcYZzU6r2lqlCLVwL6Er2beAepYjvD5ygNk5eaZnUpEREoJFaWkxNSobG92fvCE+kqJiIiI5JN+HHIyAAsEVDc7TdHxCoQ61wJws/dajp3KYs7GOHMziYhIqaGilJSYyDN9pbQDn4iIiMi/OPpJ+YWBq4e5WYpakwEADPJYC8Anf+7DMNRjVEREVJSSEuTcgU8zpURERETyS95v/1lempyf7cwSvsoZe2jidpRtcams2HvC7FQiIlIKqCglJcaxA99BzZQSERERya88Njl3OGsJ39hq2wD49M99JgYSEZHS4rKKUocOHeLw4cPOx6tXr2bMmDFMmTKlyIJJ+RN5pqeUGp2LiIhc3OrVq8nLO39D6KysLL799tsSTCTFytHkPLCGuTmKy5klfJ2ylwGwcEci+47rmlBEpKK7rKLULbfcwuLFiwGIj4/nuuuuY/Xq1Tz99NO88MILRRpQyg/HTKnUzFySM7JNTiMiIlK6RUdHc+LEP0uc/P392bt3r/NxcnIyQ4YMMSOaFAdHT6nyuHwPnEv43JN2clvt0xgGTFm69+K/JyIi5dplFaW2bNlC27ZtAfj2229p2rQpy5cvZ/r06UydOrUo80k54uXuQlU/e+POA+orJSIickH/bgRdUGNoNYsuR8rz8j3It4TvnqqbAfhmzUF2xKeaGEpERMx2WUWpnJwcPDzsxYUFCxbQr18/ABo2bEhcnLZ4lfPTDnwiIiJFx2KxmB1BioLNBimH7PfL60wpgCY3AFD96Fx6NwvFZsDzP21TcVVEpAK7rKJUkyZN+PDDD/njjz+IjY2lZ8+eABw9epTKlSsXaUApX2o4d+BTDwERERERAE7FQV42WFzAL9zsNMWnQS+wusGxHTzb1oqHq5UVe08wb2u82clERMQkl1WUeu211/joo4/o2rUrQ4YMoUWLFgD89NNPzmV9IgWJDHI0O9dMKRERkYvZtm0bmzZtYtOmTRiGwY4dO5yPt27danY8KSqOflIB1cHF1dwsxemsJXxhh+dyd+faALz0y3Yyc87f1F9ERMqvy/rW69q1K8ePHyc1NZVKlSo5nx85ciTe3t5FFk7Kn0jHTCkt3xMREbmobt265VvadP311wP2ZXuGYWj5Xnnh2HmvPC/dc2hyA+yaB9tmc89dj/Lt2sMcPnmaT/7cx6hr6pqdTkRESthlFaVOnz6NYRjOgtSBAweYNWsWjRo1IiYmpkgDSvlSQ0UpERGRQtm3b5/ZEaSklPcm52c7awmfd/IunuzdkAe/3sDkxbsZdFV1QgM8zU4oIiIl6LKW7/Xv35/PP/8csG9H3K5dO9566y0GDBjABx98UKQBpXxxNDqPT83UNG0REZELiIyMvOjt1KlTZseUopBcgYpSXoFQt5v9/tZZ9GsRTlRkJTKy83ht7g5To4mISMm7rKLUX3/9RadOnQD47rvvCAkJ4cCBA3z++ee8++67RRpQypcgH3d83F0wDDh8UrOlRERELtWpU6eYMmUKbdu2dfb1lDLOsfNeYA1zc5SUpjfaf66biiU3i/F9m2CxwKz1R1h34KS52UREpERdVlEqIyMDPz8/AObPn8/AgQOxWq20b9+eAwcOFGlAKV8sFgs1KqvZuYiIyKVaunQpw4YNIywsjDfffJNrr72WlStXmh1LikLaMftP32Bzc5SUJgMgIALSEmDDdJpVD+A/UdUBeOHnrdhsxoV/X0REyo3LKkrVrVuX2bNnc+jQIebNm0ePHj0ASExMxN/fv0gDSvnjWMKnopSIiMiFxcfH8+qrr1KvXj3+85//4O/vT1ZWFrNnz+bVV1+lTZs2ZkeUopCWYP/pG2JujpLi4gYdHrDfXzYB8nJ4JKYBvh6ubDycwvd/HTY1noiIlJzLKkqNGzeORx55hJo1a9K2bVuio6MB+6ypVq1aFWlAKX+0A5+IiMjF9e3blwYNGrBp0yYmTJjA0aNHee+998yOJUUtLwdOJ9nvV5SZUgBX3Q4+VSH5IGz5nmA/T+6/1r773mtzd3IqM8fkgCIiUhIuqyh14403cvDgQdauXcu8efOcz3fr1o133nmnyMJJ+aQd+ERERC7ut99+Y8SIETz//PP06dMHFxcXsyNJcUg/bv9pcQGvIHOzlCQ3L2h/n/3+H2+DzcYdHWtRq4oPx9Oy+GzZflPjiYhIybisohRAaGgorVq14ujRoxw+bJ9i27ZtWxo2bFhk4aR8igxy9JRKNzmJiIhI6fXnn39y6tQpoqKiaNeuHZMmTeL48eNmx5Kilp5o/+lTBayXfWleNrUZAR4BcHwn7PwFd1crY7rXA+DTZftIz8o1OaCIiBS3y/rms9lsvPDCCwQEBDi3JA4MDOTFF1/EZrMVdUYpZxzL9w6dPK1GliIiIufRvn17Pv74Y+Li4rj77rv5+uuvCQ8Px2azERsby6lTp8yOKEUh7UxRqiIt3XPwDIB2I+33/3gLDIPrm4dTs7I3yRk5zFh10Nx8IiJS7C6rKPX0008zadIkXn31VdavX8/69ev573//y3vvvcezzz5b1BmlnAkL8MTVaiE710Z8aqbZcUREREo1Hx8f7rzzTv788082b97Mww8/zKuvvkpwcDD9+vUzO55cKUdRyqcCFqUA2t0Lbt5wdD3sWYSL1cJ9Xe29pab8sZfMnDyTA4qISHG6rKLUtGnT+N///se9995L8+bNad68Offddx8ff/wxU6dOLeKIUt64ulipVskL0A58IiIil6JBgwa8/vrrHD58mK+//hqLxWJ2JLlS6RV4phSAT2WIGm6//8fbAAxoVY3wAE+Oncpi5tpD5mUTEZFi53o5v5SUlFRg76iGDRuSlJR0xaGk/KsR5M2BExkcTEonuk5ls+OIiIiUOnfeeedFx1SurO/QMi/tmP2nT1Vzc5gpejSs/hgO/AkHV+Jeoz13d6nDcz9t5cMlexnctgZuLhWs35aISAVxWf91b9GiBZMmTTrn+UmTJtG8efMrDiXlX6R24BMREbmgqVOnsnjxYpKTkzl58mSBt+TkZLNjypWq6DOlAAKqQctb7PfPzJa6uU0EVXw9OJJ8mtnrj5gYTkREitNlzZR6/fXX6dOnDwsWLCA6OhqAFStWcOjQIX799dciDSjl0z878KkoJSIiUpB7772Xr776in379nHHHXdw2223ERQUZHYsKWppCfafviHm5jBbxwdh/Rewax7EbcIzrDn/16kWr/62gw9+38PAq6rjYtVyVRGR8uayZkp16dKFv//+mxtuuIHk5GSSk5MZOHAgW7du5YsvvijqjFIO1dBMKRERkQuaPHkycXFxPPbYY/z8889ERERw0003MW/ePAxDu9eWG1q+Z1e5DjQZaL//p3221G3tIwnwcmPv8XR+2xJnYjgRESkul704Ozw8nJdffpnvv/+e77//npdeeomTJ0/yySefFGU+Kaccy/c0U0pEROT8PDw8GDJkCLGxsWzbto0mTZpw3333UbNmTdLS0syOJ0VBy/f+0Wms/efW2fDX5/i65HFHx5oATFq0W8VYEZFySB0DxRQRlexFqZTTOaRk5JicRkREpPSzWq1YLBYMwyAvL8/sOFIU8nIg48wmQT4qShHSBBr3Bwz46X54pwl3GzOJcE9jR/wpFm5PNDuhiIgUMRWlxBQ+Hq5U8fUA4EBSuslpRERESqesrCy++uorrrvuOurXr8/mzZuZNGkSBw8exNfX1+x4cqXSjwMGWFzAW/3CABjwIXR/HvyrQfoxvJa9zmKX0bzmOoWfY2M1W0pEpJxRUUpMoyV8IiIi53ffffcRFhbGq6++yvXXX8+hQ4eYOXMmvXv3xmrVJVy54Fi651MFrC7mZikt3L3h6jHw4EYY9AlUi8LVyOZm19+ZeHIUx6feCjbNFBQRKS8uafe9gQMHXvB1bUsslyIyyJt1B06q2bmIiEgBPvzwQ2rUqEHt2rVZsmQJS5YsKXDcDz/8UMLJpMg4m5xr6d45XNyg2Y3QdBAcWs3WH16h4cnfqXrgF9g6y/6aiIiUeZdUlAoICLjo60OHDr2iQFJxOHfg00wpERGRcwwdOhSLxWJ2DClOzibnFXznvQuxWKBGOwKHf8V7b93PGJeZpMx/hYAmA0EzBkVEyrxLKkp99tlnxZVDKiDn8j31lBIRETnH1KlTzY4gxS0twf5TM6UuqlqgF15X30fq8jkEnNpNwuqZhLS/2exYIiJyhfTnBTFNjSAfQDOlREREpIJyLN/zVVGqMP7vulYs8LsBgLTYV8jKyTU5kYiIXCkVpcQ0NYLsM6XiUjPJylXDShEREalgnMv3VJQqDBerhatvf5Z0PKmTt48fvvqf2ZFEROQKqSglpqni6463uwuGAYeSTpsdR0RERKRkpTl231NRqrCCQ8I53sjew7bx7o+I3RpvciIREbkSKkqJaSwWi3O21CHtwCciIiIVTbpj+Z4anV+KyD6PkWP1oIV1L7O+m0Zciv64KSJSVqkoJaZyNjs/oWbnIiIiUsE4Gp37hpibo6zxrYq1zQgA7sz7jge/Wk+ezTA5lIiIXA4VpcRUkZXtzc4PaKaUiIiIVCR5uZCRZL+v5XuXzOXqBzFcPGht/RuXg3/y3qJdZkcSEZHLoKKUmMqxfE878ImIiEiFknEcMMBiBe8gs9OUPX6hWK6y95Z6wGUW7y7cxcq9J0wOJSIil0pFKTGVc/meZkqJiIhIReJocu5dBawu5mYpq64eA1Y3ol22cRU7eOy7TWTmlLEdndMSIfY52PYTGFqCKCIVj4pSYqrIIPvyvYNJGeTk2UxOIyIiIlJCHEUpXy3du2wB1aHlLQA84vkTB5My+OTPfSaHugRH18OUrrBsAnx7O0zpAn/PV3FKRCoUU4tSS5cupW/fvoSHh2OxWJg9e3a+1w3DYNy4cYSFheHl5UX37t3ZtSv/evGkpCRuvfVW/P39CQwMZMSIEaSlpeUbs2nTJjp16oSnpycRERG8/vrr52SZOXMmDRs2xNPTk2bNmvHrr79echa5dNUreeHv6Up2ro3tcalmxxEREREpGekqShWJqx8CiwvtjQ20tOxm8uLdxKdkmp3q4jZ/B5/2hNQjEFgD3H0hbiPM+A98GgP7lpqdUESkRJhalEpPT6dFixZMnjy5wNdff/113n33XT788ENWrVqFj48PMTExZGb+80Vz6623snXrVmJjY5kzZw5Lly5l5MiRztdTU1Pp0aMHkZGRrFu3jjfeeIPx48czZcoU55jly5czZMgQRowYwfr16xkwYAADBgxgy5Ytl5RFLp3VaiEqshIAa/efNDmNiIiISAlxzJRSk/MrE1QLWgwGYLrnq7xhvMXvX70BKYdNDnYetjz7cr3vR0BuJtTrAff8CQ9ugg4PgKsXHFoF0/rCtH5wYDnk6P9viEj5ZTGM0jE/1GKxMGvWLAYMGADYZyaFh4fz8MMP88gjjwCQkpJCSEgIU6dOZfDgwWzfvp3GjRuzZs0aWrduDcDcuXPp3bs3hw8fJjw8nA8++ICnn36a+Ph43N3dAXjiiSeYPXs2O3bsAODmm28mPT2dOXPmOPO0b9+eli1b8uGHHxYqS2GkpqYSEBBASkoK/v7+RfK5lQeTF+/mjXk76dMsjMm3XmV2HBERkSum7/yC6XM5y7ynYcUk6HA/9HjJ7DRl28kD9iJO8oH8z1dtBHW7Qb3roGYn83t3ZabA9/8Hu+bbH1/9EFz7bP5cp+Jh6ZuwbirYcv553icYAiMgIOLMzxr2GVaOm4fv+c8ZvxniNtlnYiXtAc9A8AsB31DwCwXfEPvPoDrgU7m43r2IVDCF/c53LcFMl2Tfvn3Ex8fTvXt353MBAQG0a9eOFStWMHjwYFasWEFgYKCzIAXQvXt3rFYrq1at4oYbbmDFihV07tzZWZACiImJ4bXXXuPkyZNUqlSJFStWMHbs2Hznj4mJcS4nLEwWuXzOmVIHkjAMA4vFYnIiERERkWKmmVJFp1IkPLAejq4n9qfpVI5fSgvrXlyObYdj2+3Fv8p1oeOD0PxmcPUo2Xw5mXBsh70gdWIXuHpC/8nQ7MZzx/qFQp83oeMDsOR12DoLstPsyz3TE+HIuoLP4V35nwKVf3X7ssD4TZC0t/A5ra7QdiR0eQy8Kl3eexURuUSltigVHx8PQEhISL7nQ0JCnK/Fx8cTHJz/i9zV1ZWgoKB8Y2rVqnXOMRyvVapUifj4+Iue52JZCpKVlUVWVpbzcWqqeiYVpEX1QFytFhJSszh88jQRQd5mRxIRESl3li5dyhtvvMG6deuIi4vLN0MdYPjw4UybNi3f78TExDB37lzn46SkJO6//35+/vlnrFYrgwYNYuLEifj6nmeWhpxfWoL9p2/IhcdJ4VhdoHprWg1txjVv/o5L5knej06lg7ERdvwMJ3bDT/fD4lcgehREDQMPv6LNcHQ9bPnePtspLQFOJdh/Zib/M8a/GgyeDuGtLnyswBrQfxL0ew9On4Tkg5ByyP4z+dBZ9w/aj59xwn47uv7cYwVEQFgLCG0OVetD1qkz2eLPyhpvP+bK92HjV9D1SWh9J7i4FeUnJCJyjlJblCoPXnnlFZ5//nmzY5R6Xu4uNKkWwMZDyaw7cFJFKRERkWLg6OV55513MnDgwALH9OzZk88++8z52MMj/4ySW2+9lbi4OGJjY8nJyeGOO+5g5MiRzJgxo1izl0vpx+w/fauam6OcqeLrwYPd6vHSL9t5YHNlFj1yB/69XrUvh1sxGU4dhflPw9I37LOC2t0NPlWu7KSGAas+sh/XllvwGBd3qNUFBrx/ac3tLRbwDrLfwlsWPCYzxV6ochSpUg6BT1V7ISqshf13C2P3Qvuy0mPb4bfHYM3/7EtL6/Ww5xARKQaltigVGhoKQEJCAmFhYc7nExISaNmypXNMYmJivt/Lzc0lKSnJ+fuhoaEkJCTkG+N4fLExZ79+sSwFefLJJ/MtC0xNTSUiIuLCb7yCah1ZiY2Hkll7IIkBraqZHUdERKTc6dWrF7169brgGA8PD+d1z79t376duXPn5uvl+d5779G7d2/efPNNwsPDizxzuable8VmWIeafLX6IHuOpfPugl08c31je++utiNh0zewbKJ95tTS1+3FpNu+h4g2l3ey7HT46QHY8p39cf2eENnxTJ8mR9+mEHsfp+Iq7HgGQGgAhDa9suPU7WYvnP01DRb/F47/DTNugtpdIWq4/Twe/mdufvabu48KViJyRUzdfe9CatWqRWhoKAsXLnQ+l5qayqpVq4iOjgYgOjqa5ORk1q37Z231okWLsNlstGvXzjlm6dKl5OT80ygwNjaWBg0aUKlSJeeYs8/jGOM4T2GyFMTDwwN/f/98NylYa+3AJyIiYrrff/+d4OBgGjRowL333suJEyecr12sl2dBsrKySE1NzXcTIC/XvtQKLm3WjBSKm4uVcX2bADB1+X52J6bZX3D1gKuGwqjVcNPnENIUslLgy0FwdMOln+j4bvi4m70gZXGBmP/CkK/t/aBa3Gwv5gQ3tPdnKiuFGxdXaDMCHvjL3oPLxR32/g4zh8MXN8D/usHkNvB2Q3ilGvw3HFZ/bHZqESnDTC1KpaWlsWHDBjZs2ADYG4pv2LCBgwcPYrFYGDNmDC+99BI//fQTmzdvZujQoYSHhzv7HzRq1IiePXty1113sXr1apYtW8bo0aMZPHiw8691t9xyC+7u7owYMYKtW7fyzTffMHHixHwzmB588EHmzp3LW2+9xY4dOxg/fjxr165l9OjRAIXKIlcmqqa9KLUz4RSpmTkXGS0iIiJFrWfPnnz++ecsXLiQ1157jSVLltCrVy/y8vKAwvXy/LdXXnmFgIAA500zxs/IOA4YYLHaG1RLketSvyrdGwWTazN4Yc428m04bnWBxv1hxHyoEW0vTH1xAyRuL/wJtv8MU7ral7r5hsDwOfZeVWWl+HQxngFw3Qv2Al6r2+2fU3AT+65/noH2IhxATgb8+ijsWmBqXBEpu0xdvrd27VquueYa52NHoWjYsGFMnTqVxx57jPT0dEaOHElycjJXX301c+fOxdPT0/k706dPZ/To0XTr1s3ZcPPdd991vh4QEMD8+fMZNWoUUVFRVKlShXHjxjFy5EjnmA4dOjBjxgyeeeYZnnrqKerVq8fs2bNp2vSfKbCFySKXL9jPkxpB3hxMymD9wWS61Fd/BRERkZJ09m7CzZo1o3nz5tSpU4fff/+dbt26XdYx1crgPBxL97yr2AskUiye6dOYpX8fZ+nfx1i4PZHujf/VVN7dB275Fj7vD0f/gmn94I7foErd8x80NxsWvwzLJtgf1+gA//nMvmteeRRUy95w/d8Mw16Qmvukfbnf93fCyN8hqHaJRxSRss1i5PuzgRSn1NRUAgICSElJ0VK+Aoz9ZgM/rD/CA9fWZWyPBmbHERERuWyl/TvfYrGcs/teQapWrcpLL73E3XffzaeffsrDDz/MyZP/LLXPzc3F09OTmTNncsMNN1z0vKX9cykxuxfYl4yFNIV7l5mdplx79bcdfLhkD5GVvZn/UGc8XAsoAmYk2QtSCZvtu+Pd8StUqpl/THY6/PU5LH8PUo/Yn2s/Cq57vmLvUJebBVP7wOE1ENwYRsSCh3bjFJHCf+eX2p5SUvE4lvCtPaC+UiIiImY7fPgwJ06ccG7yUphenlJIaWd23vPRzPDiNvraugT7eXDgRAaf/Lmv4EHeQXD7LKjSwF5wmtYPUs4Unk6fhCWvwztNYe4T9td9Q+E/U6Hnfyt2QQrsfbpu+sK+hDFxG/w4yj6LSkSkkFSUklKjdaR9u9oNh5LJzbOZnEZERKR8uVAvz7S0NB599FFWrlzJ/v37WbhwIf3796du3brExMQAhevlKYWUdmbXZzU5L3a+Hq481bsRAJMW7SY+JfM8A6vC0B+hUi1IPgCf94P5z9iLUYtfhtNJ9teunwBjNkGTi88MrDD8w+yN461usG32P0sbRUQKQUUpKTXqBfvi7+lKRnYe2+NOmR1HRESkXFm7di2tWrWiVatWgL2XZ6tWrRg3bhwuLi5s2rSJfv36Ub9+fUaMGEFUVBR//PEHHh4ezmNMnz6dhg0b0q1bN3r37s3VV1/NlClTzHpLZVf6mZlSKkqViP4tw4mKrERGdh6v/HaBZub+YTDsJwiIgBO77Uv1stPsyywHfQKj10LrO+yzgyS/Gu2h12v2+wuety9RFREpBFMbnYuczWq1cFVkJX7feYy1B5JoVj3A7EgiIiLlRteuXblQK9F58+Zd9BhBQUHMmDGjKGNVTI5G5z4qSpUEi8XC8/2a0HfSn/y44Si3tY+kTc2gggcH1rDPmPrmdvAKhI4PQr0e5WdXveLU+k44uh7WfwHfFdD43DAgMxnSj0NAdXDzMiupiJQiKkpJqdLaWZQ6yR0da5kdR0RERKTopZ8pSmmmVIlpWi2AwW1q8NXqgzz341Z+vv9qXKznKTRVrgP3LS/ZgOWBxQJ93oLE7XBkLXxxA1Spby/Cph+z3/Ky7WO9Ktl3Poxoa25mETGdlu9JqRJ1pq/Uuv0nL/jXXBEREZEyS43OTfFIj/r4e7qyLS6Vr9ccNDtO+eTqATd/YZ8FeHI/7JoPcRvsDeIdBSkXd3sD+c/7wy4t8xOp6DRTSkqVlhGBuFotxKdmciT5NNUreZsdSURERKRoqdG5KSr7evBwjwY899NW3py3kz7Nwgj0djc7VvnjHw7D58DO3+w7G/oE2xvJ+1S13zfy4Nuh9r5TX90MN3wEzW40O7WImEQzpaRU8XJ3oUm4PwDrDpw0OY2IiIhIEcvLhYwT9vu+IeZmqYBubVeDBiF+nMzI4e3Yv82OU35VbQBXj4GrhkKDnlAtyt6vy80T3H1g8FfQ9Eaw5cL3/wertGGCSEWlopSUOo4lfGv3qyglIiIi5UzGCcAAixW8K5udpsJxdbHyXL/GAHy58gDb41JNTlRBubrDwI+h7UjAgN8ehcWv2JuhFyQ3659lryJSrqgoJaVO65qVAFirmVIiIiJS3jianHtXBquLuVkqqA51qtCnWRg2A577aSt5NvUxNYXVCr1eh65P2R8veRV+fRTit8Dm72DRS/DNbfBea3g5DN6sC3+8ZW5mESly6iklpU7rSHtRamd8Kqcyc/DzdDM5kYiIiEgRcfST8lE/KTM91acRC3cksHpfEkM/XcXEwa2o4uthdqyKx2KBro/be0/9+iis+dh+O5+FL0BuNnR9wv67IlLmaaaUlDrB/p5EBHlhM2D9wWSz44iIiIgUHccSJDU5N1W1QC/euaklXm4uLNt9gj7v/sGa/Ulmx6q42t4FN34CXkHgGQAR7eGqYdDzVbh9FozdDt3H28cueRUWvXj+pX4iUqZoppSUSq0jgziUdIS1+5PoXF/bJYuIiEg54Vi+p6KU6Xo1C6NusC/3Tv+L3YlpDJ6ykid6NuT/OtXColk4Ja/pIGh8g30GVEGf/9UPgdUN5j9tX8aXlw3XvagZUyJlnGZKSakUFam+UiIiIlIOpZ0pSvnoj26lQb0QP34c1ZH+LcPJsxm8/Ot27v5iHSmnc8yOVjFZrRcuMnUYDb3esN9f/h7MfUIzpkTKOBWlpFRyNDvfcCiZ3DybyWlEREREiki6lu+VNj4erky4uSUvDmiKu4uV+dsS6Pven2w9mmJ2NClIu5Fw/Tv2+6s+hF8eBpv+/4JIWaWilJRK9YP98PN0JSM7j+1xp8yOIyIiIlI01Oi8VLJYLNzePpLv7o2mWqAXB5MyuPmjlaxVn6nSqfWd0G8SYIG1n8DMYfYd+47vUoFKpIxRUUpKJavVwlU1HEv4dDEgIiIi5YQanZdqzasH8ssDV9O+dhBpWbkM/XQ1K/eeMDuWFOSq2+GGj8Bihe0/wfcjYFJreKU6/O86+wyqddPgVLzZSUXkAlSUklKr9Zm+Un/sOm5yEhEREZEiokbnpV6gtzufDW9Lp3pVyMjOY/hnq1m2W9ejpVKLm2HoT/aZU9Vag6sX5KTD4dWw5n/w8wPwXmtY+6lmUImUUipKSanVq1koAIt2JLI7UUv4REREpIyz5UHGmVk3Wr5Xqnm5u/Dx0NZ0bVCVzBwbd05dw5K/j5kdSwpSq5O9x9RdC+GpIzBqNQz8H3S4H0KbQfYpmPMQfN4PkvaanVZE/kVFKSm16gb70aNxCAAfLtEXiIiIiJRx6cfBsAEW8K5sdhq5CE83Fz66PYrujYLJyrVx17S1LN6RaHYsuRCrC1RtAM3/Az1egpFLoOer9hlU+/+A9zvAivftBWIRKRVUlJJS7d6udQCYvf4IR5JPm5xGRERE5Ao4lu75VAEXV3OzSKF4uLrw/q1R9GwSSnaejZFfrCV2W4LZsaSwrC7Q/l64bznU7AS5p2Hek/BpTzi20+x0IoKKUlLKtapRiejalcm1GfzvD82WEhERkTIszVGU0tK9ssTd1cp7t7SiT7MwcvIM7v1yHd+tO2x2LLkUQbXtvaeufwfc/ew9pz68Gr4dChu/hgxtrCRiFhWlpNS77xr7bKmvVx8iKT3b5DQiIiIilyndsfNeVXNzyCVzc7EycXBLBrQMJ9dm8MjMjTz23UZOZ2sZWJlhtdoboo9aCXWvg7xs2PYjzLob3qgDn/aCZe/C8V1mJxWpUFSUklLv6rpVaFrNn9M5eUxdvt/sOCIiIiKXRzOlyjRXFytv3dSSh7rXx2KBb9ceZsDkZexOTDM7mlyKgOpw60z4v0XQ6REIaWrv9XZwOcQ+C5NawwcdIWGr2UlFKgQVpaTUs1gs3Ne1LgDTlu8nLSvX5EQiIiIilyHtTC8iXxWlyioXq4UHu9dj+oh2VPH1YGfCKfpN+pPZ64+YHU0uhcUC1aOg27Nw7zJ4cBP0egNqXwNWN0jYAp/0gJ1zzU4qUu6pKCVlQkyTUGpX8SHldA5frz5odhwRERGRS+dcvqeiVFnXoW4Vfn3wajrUqUxGdh5jvtnAE99vIjNHy/nKpEqR0G4kDJ0ND++0N0XPToOvBtuX9BmG2QlFyi0VpaRMcLFauLtLbQA+/mMvWbn6whcREZEyRsv3ypVgP0++GNGOMd3rYbHA12sOMWDyMjYeSjY7mlwJn8pw+yyIugMw7Ev6fhwNuVlmJxMpl1SUkjJjQKtqhPh7kJCapSnSIiIiUvao0Xm542K1MKZ7fb48s5xvR/wpBry/jCd/2MxJbdBTdrm42Xfq6/U6WKyw4Uv4vD+kHzc7mUi5o6KUlBkeri7c1ck+W+rDJXvJs2karYiIiJQhjp5SmilV7nSsW4XfHuzEwKuqYRjw1eqDXPPW73y1+iA2XbOWTRYLtLvb3hTdIwAOroCPr4GEbWYnEylXVJSSMmVw2xoEeLmx73g687bGmx1HREREpHBseZBxwn7fN8TcLFIsqvp58PZNLfn27mgahvqRnJHDkz9s5oYPlrPpcLLZ8eRy1e0O/7cAKtWC5IMwrS+knzA7lUi5oaKUlCm+Hq4M61ATgPd/342hpoMiIiJSFmScsG87jwW8K5udRopR21pBzLn/asZd3xhfD1c2Hkqm/+RlPPbdRvYeSzM7nlyOqvXhrkVQtRFkHIe5j5udSKTcUFFKypzhHWri5ebCliOpzN+WYHYcERERkYtzNDn3rgwuruZmkWLn6mLlzqtrsejhLtzQyr6k79u1h+n29hLu+WId6w+eNDuiXCrvIBgw2d5javNM2DnX7EQi5YKKUlLmBPm4O2dLPfnDZhJTM80NJCIiInIxjn5SvuonVZEE+3vyzs0t+f7eaLo3CsYwYO7WeG54fzk3fbiChdsT1HOqLKkWBdGj7PfnPASZKebmESkH9GcaKZMeuq4eS/8+xra4VB76dgNf3NkOq9VidiwRERGRgjl33lNRqiKKigzif8OC2JVwiilL9zJ7wxFW709i9f4k6gX70rFuFUIDPAkL8CQswIuwAE9C/D1xd9UcglKn61Ow4xdI2gvzn4V+75qdSKRMU1FKyiQPVxfeu6UV17/7J8t2n+CDJXsYdU1ds2OJiIiIFMyxfE8771Vo9UL8eOM/LXi4RwM+W7aP6asOsisxjV2JBfeaqhboxa3tazC8Q0283fV/3UoFd2/oNwmm9oa/pkHTQVC7i9mpRMosld6lzKpT1Zfn+zcB4O3Yv/lLa/NFRESktNLyPTlLaIAnT/ZuxPInr+XVgc24t2sdbmhVjfa1g6hZ2RuPMzOkjiSf5vW5O+n8+mI++XMfmTl5JicXAGp2hNYj7Pd/uh+y083NI1KGqdwuZdp/oqrzx67j/LzxKA98tZ5fHuhEgJeb2bFERERE8tPyPSmAv6cbg9vWOOd5wzA4mZHD4h2JTFy4i4NJGbw4ZxsfL93LqGvrcnPrCC3tM1v38fD3PEg+AItehp7/NTuRSJmk/5JJmWaxWHj5hqZEBHlx+ORpnpq1GcNQs0gREREpZZwzpULMzSFlgsViIcjHnUFR1Vn4cBdeHdiM8ABP4lMzeXb2Fq5963e+WHmA42lZZketuDz9oe9E+/2V78OhNebmESmjVJSSMs/f0413B7fC1Wrhl01xfLv2kNmRRERERPJz9JTSTCm5RG4uVga3rcHiR7vyfL8mVPXz4PDJ0zw7ewttX17ATR+u4H9/7OVQUobZUSueet2hxRDAgB9HQa6KhCKXymJoWkmJSU1NJSAggJSUFPz9/c2OU+588PseXpu7A083K3Puv5q6wX5mRxIRkQpK3/kFq9Cfy+t1IOM43LMMQpuanUbKsNPZeUxfdYDZG46w5UhqvtcahfkT0ySE65uH6Vq4pGQkweR2kJ4IngEQWAMCakBAdQiMsP8MaQpV6pmdVKREFfY7X0WpElShL8RKgM1mMPTT1fy5+zgNQ/344b4O2qVERERMoe/8glXYzyUvB16sChjwyG7wrWp2IiknjiSfZv7WeOZvTWD1/iTybP/8X7sWEYHceFU1+rYIJ9Db3cSUFcDf82DmHZBzvobnFug/GVrdWqKxRMykolQpVGEvxEpQ4qlMek/8g+Np2fRqGsrkW67CarWYHUtERCoYfecXrMJ+Lqlx8HZDsLjAs8fA6mJ2IimHTqZns3BHIr9tjmPJ38fIPVOgcnex0q1RMDdGVadz/aq4uaiDS7HISoPkg5By6MztMCQfghO7IG4juHrBXYsgpLHZSUVKhIpSpVCFvRArYWv2J3HLxyvJyTO4/9q6PNyjgdmRRESkgtF3fsEq7OcStxE+6gy+ofDITrPTSAVwPC2LHzcc5ft1h9kW988SvyAfd5pXD6B+iB/1gn2pd+anj4dWFxQbmw2mD4I9i6BKfbhrMXj4mp1KpNgV9jtf//WRcqdNzSD+e0MzHv1uE+8t2k3dYF/6t6xmdiwRERGpqJxNzrVsT0pGFV8PRlxdixFX12Lb0VS+/+sws9cf4UR6Nr/vPMbvO4/lG18t0Ium1fzp2TSU7o1C8PN0Myl5OWS1wg1T4KNOcPxv+OVhuOFDsGg1hwioKCXl1H9aR7A7MY2Plu7l0e82USPIm1Y1KpkdS0RERCqitAT7T98Qc3NIhdQ43J/G4Y15oldDNhxKZmf8KXYlnOLvhDR2JaZxPC2LI8mnOZJ8mnlbE3B3tdKlflWubx5Gt0Yh+GoW1ZXzrQqDPoFp18Omr6FWJ2h1m9mpREoF/RdGyq3HejZkz7E0FmxPZOQX6/hxVEfCA73MjiUiIiIVjXOmlIpSYh43FyttagbRpmZQvudPpmfzd8Iplu05wZxNR9l7LJ3YbQnEbrMXqK5pUJU+zcPp1jBYy/yuRM2OcM3TsOhF+OURCL9K/aVEAHW5k3LLxWphwuBWNAz149ipLP5v2loysnPNjiUiIiIVjaMo5aPle1L6VPJxp13tyoy9rj4Lx3Zh7phO3H9tXWpV8SE718a8rQk88NV6rnoxlnu+WMfPG4+SnqVr6sty9Vio0w1yT8PMYfbm6CIVnIpSUq75erjy8dDWVPZxZ1tcKg99swGbTb39RUREpARp+Z6UERaLhYah/jzcowGLHu7Crw904r6udYis7E1Wro25W+O5/6v1RL0Uy33T1zFn01HSVKAqPKsVBk4Bv7B/+ktp3zGp4FSUknIvIsibj26Pwt3FyrytCbw5X7veiIiISAlKP9NU2jfY3Bwil8BisdA43J/Hejbk90e6Muf+q7m3ax1qBHmTmWPj183xjJ6xnlYvzOeWj1fyvz/2sjsxDW3ufhE+VeDGT8FitfeXWv+F2YlETGUx9F+NElNht0EuJb5bd5hHZm4E4JWBzRjStobJiUREpLzSd37BKuznMqmNfVbEsJ+hVmez04hcEcMw2HIklV82x/HbljgOnMjI93pEkBfXNAjm2obBdKpXFRerdpkr0B9vwcIX7PebDISuT0LV+uZmEilChf3OV1GqBFXYC7FS5O35O3l30W5crBb+N6w11zTQXyxFRKTo6Tu/YBX2c3m1BmSmwKjVULWB2WlEitS+4+ks3pHI4p2JrNqbRHaezflatUAvhrSN4OY2Najq52FiylLIZoNfHoJ1U+2PLVZoMQS6PAaVapqZTKRIqChVClXYC7FSxDAMHv52Iz+sP4K3uwvf3h1N02oBZscSEZFyRt/5BauQn0tOJrx8ppfU4/vBq5KpcUSKU3pWLiv2nGDRzkR+2xzHyYwcANxcLPRsGsZt7WrQtlYQFotmTznFb4HFL8POX+2PrW5w1VDo/Aj4h5ubTeQKqChVClXIC7FSKDvXxvDPVrN8zwmq+nkw674OVK/kbXYsEREpR/SdX7AK+bkkH4IJTcHFHZ5JBP2fcakgMnPy+HVzHF+uPMBfB5Odz9cP8WXQVdVpV7syTcL9cXNRm2MADq+DRS/C3sX2x66e0Ot1iBpmbi6Ry6SiVClUIS/ESqnUzBz+v737jo+izv84/tpN770CSei910iToohdsR4q9gaevd15ove7E8tZzoodPTsWLCiIICC9Sw+9pYf0XnZ+fwwkRKIG2OykvJ+PxzxmdmZ25rNfQuabz37Lpa8tJym9gI6R/nx+22kE+XhYHZaIiDQTeubXrUWWy6G18NZoCGwN92yxOhoRS2xOzuPDlfuZtT6Fkoqq6v0+Hm70iw9mYEIogxJC6RsXgo+nm4WRNgL7lsD8/4ODK8zE1JQ1ENzG6qhETpiSUo1Qi6yINWIpuSVc9OpS0vPLGNIulPeuH4SXewt/CIqIiFPomV+3Flku27+HT66E2H5w889WRyNiqfzSCmatT2bxjkxW78shr6Si1nF3u43hHcO5YlAco7tEttxWVIYBM86F/UugxwRztj6RJkZJqUaoRVbEGrmtKflc9vpyCssqubBPLM9f3kd93EVE5JTpmV+3Flkua2fAt3dCp/Hwl0+sjkak0XA4DHZmFLJqXzar92azel82qXml1ccjAry4pH9rrhjYhvgwPwsjtUjqr/D6SMCA6+dC3BCrIxI5IfV95rfQ1LOIqVtsIK9O7Ieb3casDSk8N2+H1SGJiIhIc1KYYa79I6yNQ6SRsdttdI4O4Ooh8bx4ZV+WPTSan+4ZyS0j2hHm50lmQRmvLdzNyGcWcuUbK/hq/SEO5RTTYtpUxPSGvleZ2z88aM7WJ9IMuVsdgIjVRnSKYNpFPXngi428tGAXCWF+TOjf2uqwREREpDkoTDfX/lHWxiHSyNlsNjpE+vPw2V2598zOzN+WzierD7J4ZybL9xxm+Z7DAAR6u9MtNpDusUF0iwmkW2wgHSL9m2dXvzGPwpZZkLoBNn4Cff5idUQiTqeklAhw2cA27DtcxKsLd/PQlxtpFeLDkHZhVoclIiIiTV11SyklpUTqy9PdzvieMYzvGcOhnGJmrjnEj1vT2ZleQH5pJSv2ZLNiT3b1+QHe7pzdI4YL+7ZicNtQ7PZmMhyHfySMuA9+mgo/PQ5dzwcvf6ujEnEqjSnlQi1yHIUmxOEwuOPj9czelEqQjwdf3X4a7SL0S19ERE6cnvl1a5Hl8vY4cxatS9+D7hdaHY1Ik1ZWWcWujEK2pOSzNSWfran5bEvJp6Cssvqc2CBvzu/Tiov6tqJzdICF0TpJZRm8Mhhy9sLwe83WUyJNQH2f+WopJXKE3W7j2ct6k5xbwoaDuVw/YzVf3j6UUD9Pq0MTERGRpkrd90Scxsvdje6xQXSPDare53AYrNqXzaz1yczelEpKXinTF+1m+qLddI0JZHyPaEZ0iqBnqyDcmmILKncvOPNf8OlEWPYy9LsGQhKsjkrEadRSyoVa5LeDTVBmQRkXvrKU5NwSBiWE8r8bB+Hl7mZ1WCIi0oTomV+3FlkuT7SC8kK4Yx2Etbc6GpFmrbSiip+3Z/DV+mR+TsqgoqrmT91gXw+GdQhnRKcIRnSMIDrI28JIT5BhwHvnwb5foNuFcNl7Vkck8qfq+8xXUsqFWmRFrInakV7AhFeXUVBWyUV9W/HcZb2x2ZrgNysiImIJPfPr1uLKpawQprUytx8+BF7NoCuRSBORW1zOD5vTWJiUwbJdh2t18QPoHBXAOb1iuLhfK1qH+FoU5QlI2wyvDwfDAdd+DwlDrY5I5A8pKdUItbiKWBO3eEcm181YTZXD4O6xnbhzbEerQxIRkSZCz/y6tbhyyd4DL/YFD1/4WwroCy4RS1RUOfj1YC6Ld2SyaGcWGw/lcuxfwUPahTKhX2vO7hmDn1cjHuHm27tg7bsQ3QtuXgh29eaQxqu+z/xmOG+miHOM6BTBPy/oDsDzP+3gm19TLI5IREREmpTCTHPtH6mElIiFPNzsDEgI5Z4zO/P15KGse+QM/nNpb05rH4bNBiv2ZHP/5xsZ8K+fuOfTDSzZmUVllcPqsI83+hHwCoS0jbD4P1ZHI+IUjTgNLGK9iYPj2ZtZxFtL9nL/zF+JC/WlT5tgq8MSERGRpuDoIOd+kdbGISK1hPh5ckn/1lzSvzXJuSV8te4QX6xLZm9WEV+uT+bL9ckE+3owtmsUZ3WPZljHcLw9GkGrJL9wOONx+O5uWPgEeHjD0DutjkrklCgpJfInHj67K3uyiliwPYOb3l/DN1OGEhPkY3VYIiIi0thVz7ynpJRIY9Uq2IcpozsyeVQH1h3I5Yt1h/hhUyo5xRV8vvYQn689hK+nG6M6RzKuRzSnd44g0NvDuoAHXA9FWfDzv2Heo2B3h8TJ1sUjcoqUlBL5E252G/+9og+XvLacpPQCbnxvDTNvTcTXU/99RERE5A8UHe2+F2VtHCLyp2w2G/3jQ+gfH8I/z+/O6n05zN2SxtwtaaTmlTJ7UyqzN6XiZrfRLy6Y4R0jGN4xnF6tg3Gzu7h77sgHoKoCFj8Nc/8Gdg8YfLNrYxBxEv1VLVIPAd4evDVpABe+spQtKfnc8+mvvDqxH3ZXP4BERESk6ahuKaWklEhT4u5mJ7F9GIntw5h6Xjc2HspjzpEE1Z7MIlbvy2H1vhyem7eDIB8PhnYIY3jHCM7uEUOQr4taUY36GzgqYMnz8MP94OZutqISaWI00LlIPbUJ9WX61f3xdLMzZ0saz83bYXVIIiIi0pgVZphr/whr4xCRk2az2ejdJpgHz+rCgntP55cHRvHERT0Z3yOaQG938koq+H5TGg9/uYlhTy3guR+TyC0ud0VgMGYqJE4xX393N6x7v+HvK+JkaiklcgIGJoTyxMU9uW/mr7z88y46RvlzQZ9WVoclIiIijZFaSok0O21CffnL4Dj+MjiOyioHG5Pz+GVHFrM3pbAjvZAXF+zinaX7uG5oAjcMa0uwr2fDBWOzwZn/AkcVrHwNvvmrOcZUn7803D1FnEwtpURO0CX9W3PryPYA3P/5RtYdyLE4IhEREWmUCjWmlEhz5u5mp19cCHeO7cicO0fw2sR+dIkOoLCskpcW7GLYUz/zzNzt5BQ1YMspmw3OmgYDbwIMmHUbzJtqjjkl0gQoKSVyEh4Y15kzukVRXung5vfXcDC72OqQREREpDExjJqWUn7qvifS3NntNsb3jOH7vw5n+lX96RoTSGFZJa/8vJthTy3gie+3kZ5f2jA3t9lg/NMw5MgsfEtfgHfPhtwDDXM/ESdSUkrkJNjtNl64vA9dYwLJKizn+hmryS/VtxEiIiJyRGkeVJWZ2/6R1sYiIi5jt9s4q0c0s+8YxutX96dbTCBF5VW8sXgPw5/6mYe/3Mi+rKKGuDGc9QRc9j54BcGhVTB9GGz7zvn3EnEiJaVETpKflzvvXDuAqEAvdmYUMvnDdVRUOawOS0RERBqDoiNd97yCwMPH2lhExOXsdhvjukcz+6/DePfagQxMCKG8ysHHqw4y+tmFTP5oHZuT85x/424XwK2LoVV/Mzn+6UT44UGoLHP+vUScQEkpkVMQE+TD25MG4uPhxi87s5j6zRYMw7A6LBEREbFa9SDn6ron0pLZbDZGdYlk5q2nMfPWREZ3icRhwOyNqZz70hKufnslczanOvfL7ZAEuG4OnHaH+XrldHj7DDi823n3EHESJaVETlGPVkG8eGVfbDb4aOUB3vplr9UhiYiIHGfx4sWcd955xMbGYrPZmDVrVq3jhmHw6KOPEhMTg4+PD2PHjmXnzp21zsnOzmbixIkEBgYSHBzMDTfcQGFhoQs/RROimfdE5DcGJoTyzrUD+eHO4ZzfOxa7DX7ZmcWtH6xj6JMLePbHJJJzS5xzM3dPc2a+v3wGPqGQ+iu8NQYOrnLO9UWcREkpESc4o1sUj5zTDYAnftjG3C1pFkckIiJSW1FREb179+aVV16p8/jTTz/Niy++yPTp01m5ciV+fn6MGzeO0tKagXknTpzIli1bmDdvHt999x2LFy/m5ptvdtVHaFqqZ97TeFIiUlvXmEBevLIvi+4fxW2ntyfc35OMgjJeWrCL4U8t4PoZq/lpazpVDif0wOg0Dm5dAq0GQEkOvHc+bP/+1K8r4iQ2Q32NXCY/P5+goCDy8vIIDAy0OhxxMsMwePTrLfxvxX68Pex8dksivVoHWx2WiIhYoLE/8202G1999RUXXnghYD7DYmNjuffee7nvvvsAyMvLIyoqihkzZnDFFVewbds2unXrxurVqxkwYAAAc+bM4eyzz+bQoUPExsb+6X0be7k41U+Pw5LnYNAtcPbTVkcjIo1YeaWDH7em8dHKAyzbfbh6f+sQHyYlJnDZwDYE+Xic4k2KYOZ1sHMu2Oxw7vPQ/9pTu6bIH6jvM18tpUScxGazMfW8bpzeOYLSCgc3vLfGec1vRUREGtDevXtJS0tj7Nix1fuCgoIYPHgwy5cvB2D58uUEBwdXJ6QAxo4di91uZ+XKlS6PudErzDDXaiklIn/C093Oub1i+eimISy4dyQ3DW9LsK8Hh3JK+Pf320icNp9/zNrM7sxT6C7t6QdXfAR9rwLDAd/eCT9PA7VREYspKSXiRO5udl66si9dogPILCjjxvfWUFRWaXVYIiIifygtzex2HhVVe/yjqKio6mNpaWlERtZOsLi7uxMaGlp9zm+VlZWRn59fa2kxio4mpTSmlIjUX7sIf/5+TjdWPDyGJy/uSeeoAIrLq/jfiv2MeXYR1767ikU7Mk9uciU3dzj/ZRjxgPl60ZNmcqpKf6+IdZSUEnGyAG8P3r52IOH+XmxLzefuTzfgcEZ/cBERkSZm2rRpBAUFVS9t2rSxOiTXqR7oXC2lROTEeXu4ccWgOObcNZyPbhzM2K6R2GywMCmTSe+s4oJXlrJge/qJJ6dsNhj9dzjnObMb37r34NOroLy4YT6IyJ9QUkqkAbQK9uGNa/rj6W7nx63pPDsvyeqQREREfld0dDQA6enptfanp6dXH4uOjiYjI6PW8crKSrKzs6vP+a2HH36YvLy86uXgwYMNEH0jpe57IuIENpuN0zqE89akgfx87+lcNzQBX083Nh7K4/oZa7jw1WUsTMo48eTUwBvgsv+Buzfs+AHeON2coU/ExZSUEmkg/eJCeGpCTwBe+Xk3s9YnWxyRiIhI3dq2bUt0dDTz58+v3pefn8/KlStJTEwEIDExkdzcXNauXVt9zoIFC3A4HAwePLjO63p5eREYGFhraREcDig6Ovueuu+JiHMkhPsx9bzu/PLAKG4Z0Q4fDzd+PZjLte+u5uLXlrH4RLv1dT0Xrvka/KMhKwneHANLXgBHVYN9BpHfUlJKpAFd1Lc1t53eHoAHvtjI+gM5FkckIiItVWFhIRs2bGDDhg2AObj5hg0bOHDgADabjbvuuot//etffPPNN2zatIlrrrmG2NjY6hn6unbtyllnncVNN93EqlWrWLp0KVOmTOGKK66o18x7LUpJDjiOjNHiF2FtLCLS7IT5e/Hw2V1Z/MAobhzWFi93O+sP5HLNO6u48JWlvL5od/0HRY8bArctgy7ngqMCfpoK718AeYca9kOIHGEzTmqENDkZLWoaZKnmcBjc8sFa5m1NJyLAi68nDyU22MfqsEREpAE1xmf+woULGTVq1HH7J02axIwZMzAMg6lTp/LGG2+Qm5vLsGHDePXVV+nUqVP1udnZ2UyZMoVvv/0Wu93OhAkTePHFF/H3969XDI2xXBpE+lZ4LRF8QuHBvVZHIyLNXEZBKdMX7uGDlfspr3RU728X4ccZ3aI4o2sUfeNCcLPbfv8ihgHr/wc/PAQVReAdBOc+Dz0muOATSHNU32e+klIu1GIqYnKcwrJKLnltGdvTCugeG8jMWxPx9XS3OiwREWkgeubXrcWUy56FZkuDiK4weYXV0YhIC5FRUMqczWnM25rOij2Hqaiq+VM/zM+Ts3pEc/nANvRsFYTN9jsJqsO74cubIPlIV+1eV5jJKU9fF3wCaU7q+8xX9z0RF/D3cuetSQMI8/NkS0o+9372q2bkExERaa40yLmIWCAywJtrEhP43w2DWfuPM3j5L325oE8sgd7uHC4q58OVBzj/5aWM/+8vvLt0LzlF5cdfJKw9XD8XRj5ozs638RN4/3woOuz6DyQtgpJSIi7SOsSX6Vf3x8PNxg+b03jmR83IJyIi0iwVHpnFUEkpEbFIoLcH5/aK5b9X9GXtP87gfzcM4vzesXi629meVsDj325l8BPzmfLROhbvyKTq2C/M3Txg1N9g0nfgHQyHVsPbZ0C2uiOL8ykpJeJCAxNCefLiXgC8tnA3H67cb3FEIiIi4nTVSSnNvCci1vNwszO8YwQvXtmX1X8by+Pnd6dbTCDlVQ6+25jKNe+s4rQn5/PE99vYkpJXM4NfwlC44UcIagPZu83EVMp6az+MNDtKSom42IT+rbl7rDlo7D9mbebn7RkWRyQiIiJOVZhprtVSSkQamSBfDyadlsD3dw7nuzuGcU1iPIHe7qTnl/HG4j2c8+ISxr2wmFcX7iI5twQiOsMN8yCqJxRlwrvnwM55Vn8MaUaUlBKxwF/HdODS/q1xGDD5o3VsOpRndUgiIiLiLEdbSvkpKSUijVePVkH884IerH5kLK9f3Z/xPaLxdLOzI72Qp+ckMfTJBVz5xgqWZ3rCdd9Du9PNmfk+uhzWf2B1+NJMKCklzUtZAcy+F2ZeB4fWWh3N77LZbDxxcU+GdwynuLyK699bzaGcYqvDEhEREWfQQOci0oR4ubsxrns0r13Vn9WPjOWpCT0Z0i4UgOV7DnPlmyu45qPtbDn9LXM2PqMKvp4MP08Dh8Pi6KWpU1JKmo+0zfDG6bD6LdjyJbw1Gv53EexfbnVkdfJws/PqxH50iQ4gs6CMa99dTV5xhdVhiYiIyKkqOpqU0phSItK0BPl4cPnAOD65OZGlD43mmsR43O02Fu/I5JxXVzG55GZy+k0xT170JHx6FZSq14ecvEadlHrsscew2Wy1li5dulQfLy0tZfLkyYSFheHv78+ECRNIT0+vdY0DBw5wzjnn4OvrS2RkJPfffz+VlZW1zlm4cCH9+vXDy8uLDh06MGPGjONieeWVV0hISMDb25vBgwezatWqBvnMchIMA9bOgLfGwOFdEBALPS8FmxvsXgDvnmX2fd79s3luIxLg7cG71w0kOtCbXRmF3PLBGsoqq6wOS0RERE5WVSUUZZnbaiklIk1Yq2Af/nlBD+bfO5IL+8Ris8HsTWkMWDGUL1o/iOHmBUmz4Y1RkL7V6nCliWrUSSmA7t27k5qaWr0sWbKk+tjdd9/Nt99+y8yZM1m0aBEpKSlcfPHF1cerqqo455xzKC8vZ9myZbz33nvMmDGDRx99tPqcvXv3cs455zBq1Cg2bNjAXXfdxY033sjcuXOrz/n000+55557mDp1KuvWraN3796MGzeOjAwNUG25sgL48ib49k6oLIWOZ8KtS2DCW/DXddD/WrB7wP4l8L8LzRkjkuY0quRUTJAP71w7EH8vd1bsyeaBzzficDSe+EREROQEFGcBBtjs4BtmdTQiIqcsPsyPF67oy/d/Hc6YLpFUOQzu3dWbi0r/QbZ7FGTvxnhrDGz+wupQpQmyGUYj+uv8Nx577DFmzZrFhg0bjjuWl5dHREQEH330EZdccgkA27dvp2vXrixfvpwhQ4bwww8/cO6555KSkkJUlNl8evr06Tz44INkZmbi6enJgw8+yOzZs9m8eXP1ta+44gpyc3OZM2cOAIMHD2bgwIG8/PLLADgcDtq0acMdd9zBQw89VO/Pk5+fT1BQEHl5eQQGBp5sschRaZth5iSzdZTNDcY8Cqf9Fey/ybXmJcOyF83WVJWl5r6oHjDsbuh+EdjdXB56XRbvyOT6GaupdBhcNzSBR8/ths1mszosERE5CXrm161FlEvqRnh9uNl1774dVkcjIuJ0q/dl8+yPSazYk00I+bzo8TLD3cy/pysG3YbHuP8DNw+LoxSr1feZ3+hbSu3cuZPY2FjatWvHxIkTOXDgAABr166loqKCsWPHVp/bpUsX4uLiWL7cHENo+fLl9OzZszohBTBu3Djy8/PZsmVL9TnHXuPoOUevUV5eztq1a2udY7fbGTt2bPU5YoFt39V01wtsZc4GMeyu4xNSAEGtYPxTcOdGM2nl6Q/pm+GLG+Cl/keSVWWu/gTHGdEpgmcu7QXAu0v38dKCXRZHJCIiIidMg5yLSDM3MCGUT25O5Ic7h3PWoB7cwt94tfJ8ADxWvca+58/gi8Vr+WTVAT5bfZDP1x7iq/WH+HpDMj8nZVBRpcHRpYa71QH8kcGDBzNjxgw6d+5Mamoqjz/+OMOHD2fz5s2kpaXh6elJcHBwrfdERUWRlpYGQFpaWq2E1NHjR4/90Tn5+fmUlJSQk5NDVVVVneds3779D+MvKyujrKwm2ZGfn1//Dy+/r7wYvrvbbPXU4Qy46HXwq0fz+IAoOPP/YPg9sOpNWPEq5Ow1u/4tfBISJ0OXcyCkLVjUQumivq3JLa7g8W+38ty8HYT4enB1YoIlsYiIiMhJKDwyvqmfklIi0rx1jQlk2sU9eeisLsxc25VHlnTjwdL/klC4Hr/5F3J/xS0sdPQ97n2tgn24aXhbLh8Yh49n4+i1ItZp1Emp8ePHV2/36tWLwYMHEx8fz2effYaPj4+FkdXPtGnTePzxx60Oo/lZ/aY5q01wPFzxEbh7ntj7fUJg5AMw5HZY9x4sexkKUuDHR8zFPxrihkBcormO6gFurvuvct3QtuQUV/Di/J08+s0WAn08uKBPK5fdX0RERE6BZt4TkRYmyNeDG4e3wzH0IVatHUXcT7cRW7aHGZ7PMC/gIj4NvpEyPKhyGOxILyA5t4THvt3KSwt2cf2wtlw1JJ4gH3X3a6kaffe9YwUHB9OpUyd27dpFdHQ05eXl5Obm1jonPT2d6OhoAKKjo4+bje/o6z87JzAwEB8fH8LDw3Fzc6vznKPX+D0PP/wweXl51cvBgwdP+DPLb5Tmw5IXzO3THzrxhNSxvPzN1lF3boDzXoTWg8xB0QvTYOssmPMgvDESnkowpzrNO3Tq8dfT3WM7MikxHsOAez/7lZ+3a1B9ERGRJkHd90SkhbLbbQwZOITY+5bD4FsBOKPgK94qe4D/nRfIRzcNYcmDo/m/C3vQOsSHw0XlPDM3iWFPLuDJH7aTWWD9kCriek0qKVVYWMju3buJiYmhf//+eHh4MH/+/OrjSUlJHDhwgMTERAASExPZtGlTrVny5s2bR2BgIN26das+59hrHD3n6DU8PT3p379/rXMcDgfz58+vPuf3eHl5ERgYWGuRU7RyOpRkQ1hH6HmZc67p7gX9J8GN8+Dhg3Dt9zD6H2bXQK9AKC+Abd/C9GGwY+6fX88JbDYbU8/rzvm9Y6l0GNz24VrW7Mt2yb1FRETkFBztvqeklIi0VB7e5pi+Ez8HvwjI2AJvnA6r3sTb3c7VQ+JZeN/pvHB5HzpHBVBQVsn0RbsZ+tQC/vbVJvZmFVn9CcSFGvXse/fddx/nnXce8fHxpKSkMHXqVDZs2MDWrVuJiIjgtttu4/vvv2fGjBkEBgZyxx13ALBs2TIAqqqq6NOnD7GxsTz99NOkpaVx9dVXc+ONN/LEE08AsHfvXnr06MHkyZO5/vrrWbBgAX/961+ZPXs248aNA+DTTz9l0qRJvP766wwaNIgXXniBzz77jO3btx831tQfaREzzjSkkhx4oTeU5cEl70CPCQ1/T0cVpP5qjmGVusHclzgFxkw9tVZa9VRR5eCm99ewMCmTAG93Pr05kW6x+tkREWns9MyvW4solxnnwr5fYMLb0PMSq6MREbFWYQbMuh12zTNfdzoLLni1ekxgh8Ng/vYMXl24i/UHcgFzeN9x3aK5ZWQ7+saFWBS4nKpmMfveoUOHuPLKK+ncuTOXXXYZYWFhrFixgoiICACef/55zj33XCZMmMCIESOIjo7myy+/rH6/m5sb3333HW5ubiQmJnLVVVdxzTXX8M9//rP6nLZt2zJ79mzmzZtH7969efbZZ3nrrbeqE1IAl19+Of/5z3949NFH6dOnDxs2bGDOnDknlJASJ1j2spmQiuwO3S5yzT3tbtCqH9zwIwy+zdy3/GV4dzzk7G/w23u42XltYn8GxIdQUFrJVW+vJCmtoMHvKyIiIidJ3fdERGr4R8LEmXDWU+DmCTvmwOvD4cBKwOzyd0a3KL687TQ+uyWRsV0jMQyYsyWNi15dxmWvL2f+tnQcjkbblkZOUaNuKdXctIhvBxtKURa80AsqiuDyD6HrudbEse07+Pp2KM0D7yAzy++CWPJKKpj41go2J+cT5ufJxzcPoVNUQIPfV0RETo6e+XVrEeXyZDyU5sLtKyGyi9XRiIg0HmmbYOa1cHgX2N1h7GNmL5TfzHy+M72ANxbvYdaGZCqqzHRFmJ8nie3DOK19OKe1DyM+zBebRTOmS/3U95mvpJQLtYiKWEOZ+3ezhVJMH7h54XG/uFwqZz98fj0krzFf97wMBt4AbQY3aFy5xeVMfGslW1LyCff35OObhtBRiSkRkUZJz/y6NftyqSyDfx1pIfXAXvANtTYeEZHGpqwAvr0TNn9hvu58Dlz4ijlD+m+k5ZXy7tK9fLTyAAVllbWOtQr2IbF9GEM7hHFGt2j8vVw3W7rUj5JSjVCzr4g1lPxUeLEPVJbCxC+g41irI4LKcpj/uJkoOyqsA/SZCL2vhMCYBrltbnE5f3lzJVtT8wn39+KTm4fQIdK/Qe4lIiInT8/8ujX7csk9CC/0MGfz/UemtV+iiYg0VoYBa96GOQ9DVTkEx8Gl75nDptShvNLBr4dyWbori2W7D7P+QE51CyoAP083LurXiomD4+ka0wyfLU2UklKNULOviDWU2ffB6jehzRC4fk7jquAdWgNr3oUtX5ldCwFsdnPmvr5XQZdzzHGpnCinqJy/vLWSban5RASYian2EUpMiYg0Jnrm163Zl8uBlfDOmRAcD3dttDoaEZHGLWW92Z0vZ5853tSZ/4JBN//p33sl5VWs2Z/N0l2H+XFLGnuOma2vf3wIVw2JY3yPGLw9nPt3mJwYJaUaoWZfEWsIuQfgxX7gqIBJ30Hb4VZHVLeyQtg6C9Z/AAeW1+zvcAZc/j/w8HHq7bKLyvnLmyvYnlZA5JHEVDslpkREGg098+vW7Mtl0+fwxQ0QPwyum211NCIijV9JLnw9GbZ/Z75uPwYueKXePU8Mw2D57sN8sHI/P25Jp/LIgOihfp5cNqAN1w1NICrQu4GClz/SLGbfE2HxM2ZCqu3IxpuQAvDyN1tGXT8HpqyFYXeDu4859ekHl5h9p50o1M+TD28cTOeoADIKyrjyzRXsySx06j1ERETkBOUdMtdBra2NQ0SkqfAJhss/gPFPg7s37J4Prw6pGXPqT9hsNk7rEM6rE/uz7KHR3HNGJ2KCvMkuKmf6ot0Me2oBD3z+K7sy9LdSY6WklDRe2Xtg/Yfm9uhHrI3lRIR3MGeSuPpL8AqE/Uvg/QugONuptwnz9+LDmwbTKcqf9PwyLn9jBTvTnZv8EhERkRNQnZRqZW0cIiJNic0Gg2+BWxabE1uV5poTS31+A5Tk1PsykYHe/HVMR355YBSvX92fgQkhVFQZfLbmEGc8v4ib31/DugP1v564hpJS0nht+hyMKmg/GtoMsjqaExd/Gkz6BnxCIXktzDgXCjOceotwfy8+umkIXaIDyCwwE1NbU/Kdeg8RERGpJ7WUEhE5eRGd4cafYMQDYHODzZ/Dq6fB7p9P6DLubnbGdY9m5q2n8cVtiZzRLQrDgB+3pnPxq8u47PXlfL0hmZyi8gb6IHIilJSSxmvPQnPd9TxLwzglsX3huu/BPwoytsA7Z5kz8zhRuL8XH980hJ6tgsguKufKN1ew8VCuU+8hIiIi9VCdlGpjbRwiIk2VmweM/jvc8COEtoeCFPjfhfD9/VBefMKX6x8fypvXDOCne0Zwaf/WeLjZWLU3mzs/2UC/f83jgpeX8NyPSazZl01llcP5n0f+lAY6d6FmP7inM5UXwZPx5nhSd6yDsPZWR3RqDu+G9y+EvANmRfWar53+mfJKKrj23VWsP5BLgJc7M64fRP/4EKfeQ0RE6kfP/Lo1+3J5Mt7sdnL7CojsanU0IiJNW3kRzHsUVr9lvg7rABe9Dq0HnPQl0/JK+d+KfczflsH2tNpDnwR4uzOsQzgX9GnF2K6RuLupDc+p0Ox7jVCzr4g5066f4IMJEBRnTqn8J9OCNgl5h8yxpQ7vAr9Ic1B0JyemCssquX7GalbtzcbX0413rh3IkHZhTr2HiIj8OT3z69asy6WsEKYdGUvqoYPg3cw+n4iIVXb9BF9PgYJUsNlh+L1mFz93z1O6bFpeKYt3ZrJ4RyZLdmWRW1xRfSw60JsrBrXhykFxmr3vJGn2PWnajnbdazeieSSkwBxf4rofIKoHFGXAh5c6ffBzfy933rtuEMM6hFNcXsW1765iyc4sp95DRERE6pCfbK69g5SQEhFxpg5j4fbl0PNSMBzmDO1vjYH0rad02eggby4b0IaX/9KPtY+cwazJQ7l1ZHvC/DxJyy/lhZ92ctqTC7jtg7Us3ZWF2vM0DCWlpHHas8hctz3dyiiczz8SrvrS7MKXvRs+vQoqy5x6Cx9PN96aNIBRnSMorXBw/YzVzFqf7NR7iIiIyG/kHRkzUuNJiYg4n08ITHgLLp1hbqdthDdGwtIXwVF1ypd3s9vo0yaYh8Z3YdnDo/nvFX0YmBBClcPgh81pTHxrJWOeXcQ7S/aSX1rx5xeUelNSShqfosPmLxmAdiOtjaUhBETBXz4Dr0DYvxS++Ss4Oevu7eHG61cP4Oye0ZRXObjr0w0892MSDoey+yIiIg1CM++JiDS87heZ4/Z1HAdV5TDvH/DueMjc4bRbeLm7cUGfVsy89TTm3DWcq4fE4+fpxp6sIv753VaGPDGfv3+1iaTfjEklJ0dJKWl89i0215HdzJZFzVFUNzPLb3ODjZ+YTVCdzNPdzstX9uPWkea4VS8u2MUdn6yntOLUv0kQERGR3zialApsZW0cIiLNXUA0/OVTOO9F8AyAgyth+jBY8jxUVTr1Vl2iA/m/C3uw8u9j+b8Le9Ax0p/i8io+XHmAcS8s5oo3lvP9plQqNHPfSXO3OgCR41R33WuGraSO1WEMnPMf+O5u+PnfENoOel7i1FvY7TYeGt+FdhF+/P2rTczemMqhnBLevKY/kQEasE9ERMRp1FJKRMR1bDboPwnaj4Zv74Td8+Gnx2Dr13DBKxDV3am38/dy5+oh8Vw1OI7lew7z/rL9zNuWzoo92azYk4273UZUoDcxQd5EB3kTG+xDTJA3MUE+JLYPI8jHw6nxNCdKSknjUz3I+elWRuEaA66Hw7th+csw6zazIhs3xOm3uWxAG+JCfbn1g7X8ejCXi15ZxluTBtA1RgOxioiIOEV1UkpjSomIuExwG7jqC9jwEcx9GFLWw+sjYcT9MPwecHNuMshms3Fa+3BOax9OSm4JH608wMerDnC4qJzk3BKSc0uOe4+fpxtXDorjhuFtiQnycWo8zYHN0BDyLtOsp0F2lpz98N9eZre2B/e1jNlrHFXw2TWw/TvwDYMbfzJbTTWAvVlFXD9jNXuzivDzdOO/V/RlbLeoBrmXiEhLpmd+3Zp1ufy3N+Tsg+vmQHyi1dGIiLQ8+akw+x5I+t58HdXTHBw9skuD3rbKYZBRUEpqXimpuaWk5pWY23klbE8tYE9WEQDudhsX9GnFLSPb0SkqoEFjagzq+8xXUsqFmnVFzFnWvQ/f3AGtB8GN86yOxnXKi+DdsyF1A/iGQ58rofeVTm92CpBbXM5tH6xj+Z7DANw0vC33j+uCp7uGmBMRcRY98+vWbMvF4YB/RYKjAu7abH5zLyIirmcYsPkL+P5+KMkGd28Y94TZQ8VmsyAcg0U7Mpm+aDcr9mRX7x/TJZKbRrRjUEIodrvr43IFJaUaoWZbEXOmz2+AzZ/DiAdg9N+tjsa1CtLMmSOy99Tsi+5lJqd6Xgr+EU67VXmlgye+38aMZfsA6N06iJeu7EdcmK/T7iEi0pLpmV+3ZlsuBWnwbGew2eGRTHDTCBkiIpYqSDeHR9k933zd5Vw4/yXwDbUspA0Hc3lj8W5+2JxWPfl6uL8nIztFMqpLBMM7RBDk23zGnlJSqhFqthUxZzEM+E9HKMqEa2dDwjCrI3K9ynLY+SP8+jHsmGt+4wpmd8aOZ8Cov0NML6fd7sctadz/+UbySioI8HLniYt7cl7vWKddX0SkpdIzv27NtlwOrYW3RkNga7hni9XRiIgImK1YV7xqDoDuqICAWLj4DWg73NKw9mYV8eYve/hmQwqFZTWzBbrZbfSLC+b0zpEMaRdGh0j/Jj1AupJSjVCzrYg5S/oWeO008PA1x5Ny97I6ImsVHYYtX5oJquS15j7vYLjhR4jo7LTbJOeWcOfH61mzPweAKwe14dFzu+Pj6ea0e4iItDR65tet2ZbLllkwcxK0GQI3zLU6GhEROVbKBvjiBji8C7CZA6Cf/rDTB0E/UeWVDtbsz2ZhUiYLkzLYkV543Dnh/l60i/CjfYQ/7Y+s+7QJJsTP04KIT4ySUo1Qs62IOcvyV80ZE9qPgau/tDqaxiVzh9n8NHmNOavPDfMgMMZpl6+scvD8Tzt4deFuDAM6Rfnz3yv6anY+EZGTpGd+3ZptuSx7GX78O/S4BC552+poRETkt8oKYc6DsP4D83Wr/nDRGxDewdq4jnEop/hIgiqTzcl5pOWX1nmem93GwIQQzugWzZndomgT2jiHYFFSqhFqthUxZ/nwMtg5F874Jwy90+poGp+iw/D2GZC925xJ4rrvnT474ZKdWdz16QayCstwt9u4fVQHpozqoEHQRUROkJ75dWu25fLDQ7DyNbP+csY/rY5GRER+z+Yv4du7oCwP3H3gzP+DATeAvfH9vVNYVsnezCJ2ZxayJ7OQ3ZlFJKUXsCujdouqLtEBnNk9mjFdImkd4kOQjwfubtZ/HiWlGqFmWxFzhqoKeCoBygvhlsUQ09vqiBqn7L1mYqooE9qdDn+ZCe7ObbqZWVDGI7M2MXdLOgCdowJ4+pJe9G4T7NT7iIg0Z3rm163ZlssnE2H7d3D2f2DQTVZHIyIifyTvEMy6HfYuMl+3GwUXvAJBrayNq54OZhczb2s6P25NY9XebBx1ZHQCvNwJ8vUg2NeDEF9PfDzccBgGVQ6DSodRve1wgL+3O+9cO9DpcSop1Qg124qYMxxYCe+cCT6hcP/uRpmpbjRS1sO750BFEfS6HC563enTmxqGwfeb0nj0680cLirHboObhrfj7jM64e2hsaZERP6Mnvl1a7bl8vpISN0AV34CncdbHY2IiPwZhwNWvwnzHoXKUvAOgrOfhZ6XOP1vq4aUXVTOgu0ZzNuaxoo92eSVVJzwNUJ8PVj/6JlOj62+z3zNVyuNw56F5rrtcCWk/kxsX7jsffjoMtj4KQTGwtjHnHoLm83GOb1iSGwfxuPfbuHrDSm8vngPP25N56kJvRjU1rqpVEVERBqdvEPmOqi1tXGIiEj92O0w+BazldRXt0DKOvjyRrPV6znPgV+Y1RHWS6ifJ5f0b80l/c3nT2WVg/zSSnKLy8ktqTDXxRUUl1fhbrdht9tws9lwd7Nht9lws9vw9rD2728lpaRxqG46ebqlYTQZHcfC+S/B17fDkufN6U0H3eT0rH6onyf/vaIv5/WK5e+zNrE3q4jLXl/O2T2juffMzrSP8Hfq/URERJqcihIozjK3lZQSEWlaIjqZk0j98iwsegq2zoK9i2Hcv6H3lU2q1RSAu5udUD9PQpvA7HxHqfueCzXbJuunqrwInowHRwXcsQ7C2lsdUdOx6Bn4+V9HXtjA3RvcvY5fe/qDpy94+IKnX8267QjoMLZev2zzSip48odtfLL6IIZhzvpw2YDW3DmmE9FB3g37OUVEmhg98+vWLMslaxe83N981j58qMn9ASMiIkckr4Ovp0DGFvN1wnA494VGNUNfU6Lue9J07F9uJqSC4iC0ndXRNC0j7oOSHFjxKmBAZYm51NeyFyGmD4x80BwD4w8q0kE+Hky7uBfXntaWZ+Zu56dtGXy86iBfrkvm2qEJ3D6yA0G+Hqf8kURERJqUvIPmOqi1ElIiIk1Zq35wyyJY/jIsfAr2/QKvJcLw+2DYXeaX/eJ0ainlQs3y20Fn+PERWPYS9L3KnPVATlxpPlQUQ2XZkaXUXFeVmd0KyovM4+VFNdsFabBpprkNEN0TRjwAXc6t17hea/Zl89Sc7azelwNAoLc7t4xszzWJ8QR4KzklIi2bnvl1a5blsv4D+Hqy2fL4qi+sjkZERJwhey/Mvhd2zzdfh3cyW00lDLU0rKZELaWk6age5Px0C4No4rwDzeVEjXnU/CZg1ZuQtgk+uxoiu5stsDqeCV6/P2bUgIRQPrslkQXbM3h6ThJJ6QU8MzeJNxbv4bqhCVx3Wlu1nBIRkebv6CDngU1jKnEREamH0LbmFw2bv4A5D0PWDphxNvS8DM543JxsSpxC05yJtYoOm8kQgHYjrY2lJfILN2fuu2sTjLgfvALNPtSfXwfTWsPLA+HLm2H5q7B/GZQVmu8zDCgrwJZ7gDFBKXx/bgUzh6ZwQch+CktKeeGnnQx7agH/mZtETlF53fc2DHMcjgMrwFHlso8sIiLiVNXd99pYG4eIiDiXzQY9L4Epq6D/tYANNn0GLw0wJ5uqLLM6wmZB3fdcqFk2WT9VW78xW+dEdIHJK62ORkpyYMV0WP8/yE+u4wQb+IZBaZ45DlgdKjwCWG704tuS7iyq6k2hZzhXJ8Zzw9C2RHqWm7NZ7J4Pu36C3APmm8I7w8gHoPtFYHdruM8nIuIieubXrVmWy/sXmK2+L5wOfa60OhoREWkoyevghwfg0GrzdWh7OOtJ6HSmtXE1Uuq+J03D/qXmOmGYtXGIyScERj1sLoUZkLIBUjdAynpzuyClZtprADcv8A0Fn1DwDoLMbXiU5DCCpYzwWAoesNURz6qlndm3/ABh9p24cUyrKLuHOUNgVhJ8cQMsfsYcdL3bhfUa10pERMRyR7vvBbW2Ng4REWlYrfrB9T/Cxk9h3qOQvRs+uhQ6joOzpmkW+ZOkpJRYa98Scx2vAeMaHf9IM+t/bOa/IB0K083klW8oePjWnmnIUWV+g7BrHuych5Gynm72/XSz768+Za8jis2+AwnueRYDTz8fbzdg5evm2FaZ282ugxFPw+kPQtcLlJwSEZHGyzCUlBIRaUnsdrNVbJdzYPHTsOI12DnXbDE76mFIvAPclGY5Eeq+50LNssn6qSjOhqfbAQbct9NMgkjzUpQFuxfAodWkeMTxfkZ73t1mo6zSAUCIrweXD4xjQr9WdAysgpXTzfGryvLM9/uGm4MMBsdDSDwEx9VsB8SAh4+FH05E5PfpmV+3ZlcuRVnwTHvABo+ka7pwEZGWJnOH2aVvz8/m65je5ozy0T2tjasRqO8zX0kpF2p2FbFTte07+HSiOb3mlNVWRyMuklNUzierD/LBiv0k55ZU7+8WE8gFfWK5oIsf0VvfhRWvQln+H1/MK8hMZvpHQUCUufYOgpJcKD58ZMkyE6DFh6Gq3Gzl5RMC3sE12z4hENUdOozRTBoi4hR65tet2ZVLynp443Twj4b7kqyORkRErGAYsOEjmPuwOfau3R2G3W1OJNWCv6zQmFLS+Gk8qRYpxM+T205vz80j2vHTtnRmrjnIwqRMtqbmszU1nyfnwKCE05kw8iLGRxcQUJoCOfshd/+R9QFzqSwxW1SV5cHhnfUPoCjTXH5P5JHkVIcxEJfYoh8kIiLyJ9R1T0REbDboOxE6jIXv74Vt35pj5W79xmw11Wag1RE2akpKiXX2/WKulZRqkdzsNsZ1j2Zc92hyisr5fnMqX69PYdW+bFbuNZe/2W0kto/mrB59OHNENBEBRxJEhmF+C1GYYY5xVZhes12aBz7B5iyBvmFmF0DfMHMMLDcPsxVVSY65lB7ZLsqE/csheS1kbDGXZS+aY2bFD4WAaPD0B08/8PI/su0PwW3M48eOqyUiIi1H3pGZapWUEhGRgCi4/APYMgu+v8+czOntM2DgjTDqb+bfI3Icdd9zoWbXZP1UlOTAU20BA+7dYf4HFgGSc0v4ZkMKX29IZntaQfV+mw0GxodyVo9ozuoRTWxwA4wnVZxt9gffNR92/WQmuf5MqwEw5lFoN9L58YhIk6Vnft2aXbnM/bs5UUfiFBj3b6ujERGRxqI4G+b+DX792HztEwKj/g79r2sxA6FrTKlGqNlVxE7F9u/hkyshrCPcscbqaKSR2pNZyJwtaczdnMavh/JqHesSHcDwjuEM7RDO4LZh+Hi6OffmhgHpm+HACrP1VXkRlBea67ICc/vACqgoNs9vOxLGTIXW/Z0bh4g0SXrm163Zlctnk2DrLDjrSRhym9XRiIhIY7N3MfzwIGRsNV9HdofxT0Hb4dbG5QJKSjVCza4idirm/A1WvAL9r4Xz/mt1NNIEJOeWMHdzGnM2p7F6fzbH/ubydLPTPz6EYUeSVN1jA/Fwszd8UIUZ8MuzsOYdcxB1gC7nwuhHILJrw99fRBotPfPr1uzK5c0xkLzG7K7R9TyroxERkcaoqhLWvgs//9vsMQTQ7QI481/m7OLNlJJSjVCzq4iditdHQOqvMOFt6HmJ1dFIE3O4sIxluw+zZGcWS3Zl1ZrFD8DHw43ebYLoHx9C//gQ+rYJIcTPs+ECyj0AC580m+caDsBmzuYH5utjF5sd4oZA94shYXiLab4r0tLomV+3Zlcu/+kMhWlw80KI7Wt1NCIi0pgVZ8PPT8Cat82/C9y94bQ7YOhd5ri1zYySUo1Qs6uInaySXHgqATDgnu0QGGNxQNKUGYbBvsPFLNmZyZJdWazYk01eScVx57WP8KNX62DahvtVLwnhfvh7OTEplJkEC/4F276p3/l+EdDtQugxAdoMBrsLWneJiEvomV+3ZlUulWXwr0hz+/7d4BdubTwiItI0pG8xu/QdnfjLPxrGToVeVzSrvweUlGqEmlVF7FQkzYGPL4fQ9vDXdVZHI82Mw2GwJ6uQtftzWLMvh7UHctiTWfS750cEeNE2zI/O0QH0bBVEj1ZBdIzyP7Xuf5k7IO+A2Srqt0tZISR9D1u/hpLsmvcEtoKOZ4KHjzmeFUbNGsyZnbpdACEJJx+XiLiMnvl1a1blkr0XXuxjftP99zTNxCoiIvVnGLD9O/jxEcjZZ+6L7QvjpkF8oqWhOYuSUo1Qs6qInYqjM9X0uwbOf8nqaKQFyC4qZ93+HJLSC9ibVcTerCL2ZRVxuKi8zvM93e10jQmkV6sgerYKoltsIB2j/PFyd+Jg6lUVsGcRbP7CfCCV5dfvfa36my2rul8EgbHOi0dEnErP/Lo1q3LZ+wu8dy6EdYA71lodjYiINEWVZbByOix6BsqPzDze/SIYcT+EtAVPX2vjOwVKSjVCzaoidireOB1S1sPFb0Kvy6yORlqwvJIK9h1JUm1JyWNTch5bkvMpKKs87lx3u432Ef50iw2ka0wA3WKC6BoTQJi/16kHUlEKu36C5LWYLaNsx3zjfmR9aLXZxNdw1OyPP818aLXqB/5R4BcJ7g04dpaI1Jue+XVrVuWy4WOYdSu0Ox2u+drqaEREpCkrzDAHQl/7HtU9JQC8giAg+sgSY66jukObQRAc36hb6db3ma8RdsW1SvPMAc4B4odaG4u0eEE+HvRuE0zvNsFc2LcVYHb/259dzKbkPDYdymVzcj7b0vLJLa4gKb2ApPQCvlpfc42IAC+6RAfQJTqAztGBdIkOoEOkP94eJ9CqysMbup5rLn+kIN3s9rf5Czi4AvYvNZdj+YSYCSr/SAiIhdYDzP9rEV2aVR91ERHL5R0y14GtrY1DRESaPv9Ic1b6gTfCvKlwYDlUFENZnrlkJdXxnihzXNqjS0zvJvkFtZJS4loHVpotPULaQlArq6MROY7dbqseCP383mb3OMMwSM0rZVtqPltTzCTV1pR89h0uJrOgjMyCMn7ZmVV9DTe7jfhQX9pF+NE+wp/2Ef7V26c0C2BAFAy+2VzyDsGWr2D7bHP2v8IMcFSY08yW5EDmdvM9Gz8x1z4hEHea2boq/jSI7qWZ/0RETkX+kaRUkJJSIiLiJNE94eovzTGnygqgIA0KUmvW+clmr6OUDVCYbk6wdHSSJTdPCOsIEZ0gvDOEd4SIzmY3c3dvc/a/3H2Qs9/8+yH3yNrNE6782LKPrL9IxLWOzjCQoFZS0nTYbDZig32IDfZhTNeo6v1FZZXsSC8gKa2A7WkFbE/LZ3taAbnFFezJKmJPVhE/bcuoda0QXw/aRfjTLtzPXEf40T7Cj7hQPzzdT6AlU1BrcwrZ0+4wXzscUJprJqcK08119h7zW5aDK81EVdJscwHz4RMcD6FtzSTxsevQ9kpYiYj8mTwlpUREpIHYbOAdaC4RnY4/XlFiJqYOroCDq8z6fvFhyNhiLrUvZk6mVFFc9708/MwkmEVdAfVXh7jWviXmOmG4tXGIOIGflzt940LoGxdSvc8wDDIKytiVUciezEJ2ZxaxO7OQPZlFJOeWkFNcwdr9Oazdn1PrWm52G21CfIgL8yM+1Jf4MF/iw/yID/MlLtT3z7sD2u3gG2oukV1qH6uqMLvN7l8K+5bCgRVmM+DDO83lt7yCoP0o6HgGdBhr9l0XEZHalJQSERGrePiYs/QdnanPMMxZ/LJ2QGaS2d0vc4e5Ls2rSUgFxEBwnPnldHAchMSb20pKSYtQmq/xpKTZs9lsRAV6ExXozdAO4bWOFZdXsifTbEG1N7OIPVlmsmpPZiFF5VXsO1zMvsN1f4MRHehNfJgvbcP9SAj3IyHMl4RwP+JD/fDx/JOElZuHObZU6wEw9E5wVEHeQXM685y9x6z3ma2ryvJg6yxzAbOrX8czoOOZ0HqQxqYSETGMY5JSbayNRURExGY70uOhLXQaV7PfMKAo0+wKGNjKHMu2kVFSSlzn4Eowqo5kZVWBk5bH19OdHq2C6NEqqNb+o62r9mQWcSC7iP2Hi9mfXcyBw8XsO1xEQWklafmlpOWXsnJv9nHXjQzwonWID21Cfc11iC+tQ8wWVrHB3ri7/SaJZHeDkARzYVTtY44qSF4Hu+bBzh/NPutpG83ll2chqgcMvxe6XWBeR0SkJSrNhfJCc1tjZIqISGNls5mDqPtHWh3J71JSSlxHXfdE6nRs66rE9mG1jhmGQW5xBfsOm8mqvVlF7DtcxL6sIvZmFZFfWklGQRkZBWWsO5B73LU93GzEh5kDt7eL8KsZyyrcj1A/T2y/baZrd4M2A81l1N+gMBN2zzcTVDt+hPTN8Pl1EN7JTE71uETjT4lIy3O0lZRvuNmFQkRERE6K/pIQ16lOSqnrnkh92Ww2Qvw8CfHzrDV21VE5ReUczCnmYHYJh3KKf7NdQnmlg10ZhezKKDzuvYHe7tWDrR9NVh2debB6DCv/COh9hbmU5MDK12HFq2Z/9a9ugYVPwvB7oNcVTXIKWhGRk6LxpERERJxCSSlxjbJCsxsQaDwpESc6mrDq1Tr4uGMOh0Fybgl7j7Sq2pNZaM4KmFlESl4J+aWVbDiYy4aDubXeZ7NB6xAfOkT40z7Cn/aR5rpDpD+hpz8EQ26H1W/C8lfMsai+uQPmPQoBseZA6z4h5uIbCj6hENvHbCFp0eCJIlI/jz32GI8//nitfZ07d2b79u0AlJaWcu+99/LJJ59QVlbGuHHjePXVV4mKiqrrcs2bklIiIiJOoaSUuMbBFeZ4UkFHRvgXkQZnt9toE+pLm1BfRnSKqHWstKKKfYeLqgdaPzoA+57MQvJLKzmYXcLB7BJ+Tsqs9b5gXw/ahfvRPmIsnfqPY3jeN3TY+Q7uJZlmS6rfE9EFBt8CvS4HT7+G+Lgi4gTdu3fnp59+qn7t7l5TVbz77ruZPXs2M2fOJCgoiClTpnDxxRezdOlSK0K1Vt5Bc62klIiIyClRUkpcY9+RCmvCMGvjEBEAvD3c6BIdSJfowFr7DcMgq7Cc3ZmF5pJRxO5Ms/tfcm4JucUVrDuQe8z4Vb3x5D90dUume0gV3YIrae9XRhvvUiLci/EqzYSd8yBzO3x3N/z0GPS7BgbepAS1SCPk7u5OdHT0cfvz8vJ4++23+eijjxg9ejQA7777Ll27dmXFihUMGTLE1aFaKy/ZXCspJSIickqUlBLX2L3AXGs8KZFGzWazERHgRUSAF0Pa1R50vaS8yuwGmGW2rNp9tIVVZiG/lifwaxaQVft6caG+9I29lottCxmY8Tm+RQdh2Utm17/OZ0PHMyCyG0R0Bu/asxKKiOvt3LmT2NhYvL29SUxMZNq0acTFxbF27VoqKioYO3Zs9bldunQhLi6O5cuX/25SqqysjLKysurX+fn5Df4ZXELd90RERJxCSSlpeGmbIXUD2N2h45lWRyMiJ8nH041usYF0iz2+dVVybglbU/LZmprPlpR8tqbkk5xbwoHsYg5kw9cMxE5/RtnXc4PHj5zGJtj+nbkcFdgaIrtAZFeI6QNdzwN3L9d+SJEWbPDgwcyYMYPOnTuTmprK448/zvDhw9m8eTNpaWl4enoSHBxc6z1RUVGkpaX97jWnTZt23DhVTZ6jCtK3mNuh7a2NRUREpIlTUkoa3rr3zHXns8E/0tpYRMTpbDYbrUN8aR3iy5nda7r95BaXszUln21pBWxLzWd7Wj6/pA9kfll/OtoOMcHtF7rZ9tHRnkyMLRvyD5nLLnM8m8rAeNzOfBxb9ws1SLqIC4wfP756u1evXgwePJj4+Hg+++wzfHx8TuqaDz/8MPfcc0/16/z8fNq0aXPKsVoqbROUF4BXIER1tzoaERGRJk1JKWlYFSWw8VNzu/8ka2MREZcK9vXktA7hnNYhvHpfZZWDfYeL2JZawLbUkcw4krAqysuioy2ZzvZDdLIdZLzbKqLy98Pn17Lzm26s6XQPEd1G0KtNEJEB3hZ+KpGWIzg4mE6dOrFr1y7OOOMMysvLyc3NrdVaKj09vc4xqI7y8vLCy6uZtXjcv8xcxw0Bu5u1sYiIiDRxSkpJw9r6NZTmmbPutRttdTQiYjF3NzsdIgPoEBnAeb1jq/fnFpezPa2A7an5bEst4NuUdEZkfMxNbt/RsXwrHTffyHe/DuaSyisoD4inR6tAuscG0aNVED1aBRId6I1NralEnKqwsJDdu3dz9dVX079/fzw8PJg/fz4TJkwAICkpiQMHDpCYmGhxpC62/8jkLfGnWRuHiIhIM6CklDSstUe67vW7Gux2a2MRkUYr2NeTIe3Cag2uXloxml17duG1eBrtk2dxrttKzrSv4aPiMfxv+xn8tK1V9blhfp50bxVEj9hAM1EVG0SbUB8lqkROwH333cd5551HfHw8KSkpTJ06FTc3N6688kqCgoK44YYbuOeeewgNDSUwMJA77riDxMTEljXznmHUtJSK1+QtIiIip0pJKWk4mTvgwDKw2aHPRKujEZEmxtvDjR6dO0PnGeaECfMexXP3fK51/5Fr3X9kl29fZjKW93N7crgIFu/IZPGOzOr3B3q7H2lNZSaqusYE0i7cD3c3JchF6nLo0CGuvPJKDh8+TEREBMOGDWPFihVEREQA8Pzzz2O325kwYQJlZWWMGzeOV1991eKoXSwzCUqywd3HnJBBRERETonNMAzD6iBaivz8fIKCgsjLyyMwMPDP39DUzf07LH8ZOp0Ff/nU6mhEpDnY/TOsfB12zgXDAYDhG0ZG+0tZFnIeq3ID2ZycT1JaAeVVjuPe7ulup3NUAF1jAugaE2gu0YEE+Xq4+pNIM9finvn11OTLZfXbMPseaDsCJn1rdTQiIiKNVn2f+WopJQ2jsgx+/djc7qcBzkXESdqPMpe8Q7DufVj3PraCVKI2TecipnNReGeI6kplj26keCawuaI1K3P92ZRSyPa0AorLq9iUnMem5Lxal40K9KJTVAAdIwPoHO1Px6gAOkb6E+CtZJWIHENd90RERJxKSSlpGNtnQ/FhCIiBjmdaHY2INDdBrWHU32DEA2arqTXvwK75kJUEWUm4M4s4IA44290HIrtijBxNavRofq1KYFtaIVtTzZn/knNLSM8vIz2/jF92ZtW6TatgHzpHB5hLlLluH+GPp7u6AIq0OLXGk9Ig5yIiIs6gpJQ0jHVHBjjvMxHc9GMmIg3EzR26nGMuBemQtgkytkDGNkjfYo7/UlkCKeuwpawjlv8QGxDD+M7j4bSzoe0ICirt7MwoZGd6ATvSC9mRXsCO9ALS88tIzi0hObeEBdszqm/pbrfRLsKPTlFmoqrTkYRVm1Bf3OwaWF2k2crZBwUpYPeAVgOsjkZERKRZULZAnC97L+xZaG73u9rSUESkBQmIMpeOY2v2OarM30mHVkPS92ZrqoJUs2XVmnfA05+AuCH0C4yln38UREZBuyjwjyLfPZ7tJUEkZZaQlGaOU7U9rYCC0sojyatCviO1+lbeHnY6Rprd/tpF+NE23J+24X4khPvi66nHrUiTd7SVVKt+4OlrbSwiIiLNhGrJ4nzr/2eu242CkARLQxGRFs7uBuEdzKXPlVBRCvt+MRNUST+YCapdP9X51kBgUEAMg/pfB2OuhYCeGIZBal4pSWlma6qkI62qdqYXUlrhqHO8KoCYIG/ahvvRNtyPdhH+tI/wo32EP7HBPmpdJdJUqOueiIiI0ykpJc5VVQnrPzS3+2uAcxFpZDy8oeMZ5nL2s5C6HlI3QlEmFKRBYToUZpjrgjQzabXwCVj8DHS/ENugm4ltPZDYYB9GdYmsvmyVw+BAdjFJaQXsTC9g7+Ei9mYVsSeziLySClLzSknNK2XZ7sO1wvF0t9Mu3I92EX7Eh/kRH+pLXKgvcWG+xAQpYSVyQooOw/bvoN81YGuA/zv7l5prDXIuIiLiNEpKiXPtnAuFaeAbDp3PsToaEZHfZ7dDq/7mUpfKMtj6Dax6Aw6tgk0zzSW6Fwy6GXpcDJ5+ALjZbdUtoc7qEV3rMjlF5ezJKmJPZmF1ompPViH7soopr3Sw/Ui3wN/ycLPROsSXNqG+tAnxObLtQ5sj+0J8PbA1xB/eIk2RYcDsu2Hr1+ZywcsQGOu86+enQM5esNmhzSDnXVdERKSFU1JKnGvd++a6z5Xg7mltLCIip8LdC3pdai4p62HVW2ZSKm0jfDMFfnjQHGC912Vmd+XfmdQhxM+T/n6e9I8PqbW/ymFwKKeYPZlF7M4s5EB2sbkcLuZgTjEVVQZ7s8wWV3Xx83SjVYgPscFHliDv6u1WwT5EBXprlkBpWeISYcdc2D0fXh0CZ/8Hel7qnFZTR7vuRfcE76BTv56IiIgAYDMMw7A6iJYiPz+foKAg8vLyCAwMdO7FDaNhmqqfiLxkeKEHGA6YsgbCO1obj4iIsxVnm+PmrXnXbDVxlF8EdL8Yel1uDoJ8ir+PqxwGafmlHDhczIHsIg7llHAwu9hc5xSTnl/2p9ew2SA60JvWIWaSqnWIL61CfKpfxwb74O3hdkpxyu9r0Gd+E9bg5ZK5A766BVLWma+7ng/nPg9+4ad23e/ugTVvw5Db4axppx6niIhIM1ffZ76SUi7UYBUxw4BnOpjf3IXEQ3B8zfrotm9Ywyatyovg06vNbyfjh8J13zfcvURErGYYkLwWNn4Km7+E4qyaY34R4OZlJuiPLhjmOjAW2o6EdqebrTq8/E/q9qUVVSTnlpByZEnOLSUlt4TUvBJScktJzi2hvNLxp9cJ9/eiVbA3rY4kqmKCfAjz9yTC34swfy/C/D0J8fXU2FYnQUmpurmkXKoqYcnzsOhJcFSa/yfPexG6nH3y13xlCGRug8s/gK7nOS9WERGRZkpJqUaowSpixdnwdNs/Pic4DtqPgfajod1I5zY9LzoMH11q/oHm7gNXf6mZaUSk5aiqgN0/w6bPYPtsqCiu3/vs7tB6kPk7ue1IiOxq/m52whcIDodBVlEZyTklHMopITm35Mh2cfXr4vKq+oVpg1A/T8L9vYgIOGY55nWgtweB3h4EeLsT4O2Ou5u6DSopVTeXlkvKBvjqVjOZBOYA6Oe+YM7KeSKKDsMz7czt+3efeqsrERGRFkBJqUaowSpiDgfkHYTc/ZB7AHL2m9s5R14XpALH/DPb3KD1ADNJ1WEMxPYzB/w9GTn74YOL4fAu8AmBv8yENgOd8rFERJqcskLI2mFu2+xHFpu5BsjYBnsXwZ6F5u/n3/L0h8BWENTKXP92OzAWvE/9+WEYBnklFbUSVim5JaTml3K4sIzDheUcLionp7ick6kl+Hi44e/tTpCPB1GBXkQFehMd6E10kHf1dqifJ4HeHvh5uTXLJJaSUnVzeblUlMLP/4ZlLwEGjH4ERtx/YtfY9h18OhEiusDklQ0SpoiISHOjpFQjZFkFtazQnMZ49wLYNR8O76x9PLA19JxgDgYa1aP+39KnbYYPJpiz7QW1gau+hIhOzo9fRKQ5yt5rJqf2LoJ9S6Eoo37v8wo0k1OBseYfyV3OMbsCnmjrj3qorHKQXVxOVkE5WYVlZBaUkVlYRka+uc4sKCWzoIz80koKSyspqahf66vfOprECvByr05mBft6EuJrrkN9PQjx8yTY15MAb3cCvd0JONI6y9vD3ihnIVRSqm6Wlcv6D+Hr280v5q6fc2Iz6M35G6x4BQZcb45PJSIiIn9KSalGqNFUUHMPmF1Nds+HXQug/JipyMM7m8mpnhMgtN3vX2PfEvj4SijLh8hucNUXzp16WUSkpSkvNqedzz9kThyRnwx5h8x1foq5Ls2r+72+4WZyqtv5kDDCstlPK6ocFJZWUlBaSX5pBbnFFaTnl5KWX0rGkXVafhnpeaXkFJdTVo9xr/6Mu912pNug2X3Q36tm++hrP69j1274ebnj62nuiwjwItTP+eXVaJ75jYxl5WIY8MWNsPlzc0iDW5fUfyiD10dC6gaY8Db0vKRBwxQREWkulJRqhBplBbWiBHb+aE5zvuNHqDpmRqfwThAQA/6R4BcJ/hHmurwIfnzEPDfuNLjyY/AJtuwjiIi0GGWFZpfso8mq/cvMcaxKc2vO8Q6CTuPNliAh8RDS1mzNalGi6o+UVzooKquksMxMZJlrM5mVU1xevc4pLienyNwuKDXPKSyrxOGEGsx1QxOYel73U7/QbzTKZ34jYGm5lObB9GHml3M9LoEJb/156/DSfHgq3pyo4O6tZndaERER+VP1fea7uzAmaYw8fKDbBeZSmmeOm7BpptmdJGtHzdgodelyrlmh8/BxXbwiIi2Zlz94dYTwjubrvleZA63v+wW2fWv+Di/KgI2fmEs1m9maNSTBTFB5+pozBLp71l57B5rnhLSFoNYN0iXwWJ7udjzdPQk5iZZKhmFQVF5FfklFdaKqoKyyuqVWYdnR/Wayq7i8ksKyKorKKs2lvJKisiqCfRpfsk4aiHcQTHgH3hlntpjqMAb6/OWP33NwlZmQCklQQkpERKQBKCklNbyDoO9EcynMgLRNUJRpbhdl1myXZJuDpI/6W4P/wSIiIn/CzcOcWbX9aDj7P+Yf0UnfQ9ZOyNlnTnxRUXykG2By/a9r9zBbWoW2M5NU4R3N7tqRXcE3tME+Tn3ZbDb8j3TLE6m3NgNh1MOw4F8w+z5zBszwDr9//v6l5jp+qGviExERaWFUk5O6+Uea3yCKiEjTYXeD+ERzOcowoCirJkGVdwgqS6GyDKrKj6zLoLLc/NIhe695XlW5ObPq4V3H38c/2kxORXY70mrLMLsWlhceWReYXb0Nhzl+z9HEVmhbcxZBfaEhVhp2D+xZZLYw/OIGuGHe73dv3b/MXMef5rr4REREWhAlpURERJozm80cE9A/wmwlUh+OKrNVVfZeyNkLh3eb3bkztprj8RSmmcuen088HjdPCI6HwBhzJkHvIHPtFWB2H/QKMLskVpSYLbzKi46si8GogoBoc9bYoGMWn5D6zxwrYneDi16H6UPNAcwX/BPO/Nfx51WUQPJac1tJKRERkQahpJSIiIjUZnczWzgFxwEjax8rK4DMJDNBlb4VsnebXf28/MHT/8g6wFwbDsjZD9l7zCX3wJEWWDvNxVk8fM0xs/yjzZa+/lG114Zhtt4qKzSTXOWF5ucoLzS7PXY9z3mxSNMQ1ArOfxk+nQjLXgKfUPNnyM3DTJzaPcyErKPCnPQlpK3VEYuIiDRLSkqJiIhI/XkFQOsB5nKiqioh/5DZAqso05xgoyzfnOGsrMDcLisAuzt4+pnJpuq1L9jskJ8KeQfNllx5h8zrVBT/flfDP+Phq6RUS9X1XBhwA6x5G+Y//vvnxZ+mlngiIiINREkpERERcQ039yOz+yU475oVJZCXDAWpUJhuTsjx27XNZibTqlty+de8jhvsvFik6Rn3b7PbaMY2sxVfVcWR5ci2mzsMvs3qKEVERJotJaVERESk6fLwMWdP+6MZ1ER+j4cPjH3M6ihERERaLLvVAYiIiIiIiIiISMujpJSIiIiIiIiIiLicklIiIiIiIiIiIuJySkqJiIiIiIiIiIjLKSklIiIiIiIiIiIup6SUiIiIiIiIiIi4nJJSJ+iVV14hISEBb29vBg8ezKpVq6wOSURERERERESkyVFS6gR8+umn3HPPPUydOpV169bRu3dvxo0bR0ZGhtWhiYiIiIiIiIg0KUpKnYDnnnuOm266ieuuu45u3boxffp0fH19eeedd6wOTURERERERESkSVFSqp7Ky8tZu3YtY8eOrd5nt9sZO3Ysy5cvr/M9ZWVl5Ofn11pERERERERERERJqXrLysqiqqqKqKioWvujoqJIS0ur8z3Tpk0jKCioemnTpo0rQhURERERERERafSUlGpADz/8MHl5edXLwYMHrQ5JRERERERERKRRcLc6gKYiPDwcNzc30tPTa+1PT08nOjq6zvd4eXnh5eXlivBERERERERERJoUtZSqJ09PT/r378/8+fOr9zkcDubPn09iYqKFkYmIiIiIiIiIND1qKXUC7rnnHiZNmsSAAQMYNGgQL7zwAkVFRVx33XVWhyYiIiIiIiIi0qQoKXUCLr/8cjIzM3n00UdJS0ujT58+zJkz57jBz0VERERERERE5I8pKXWCpkyZwpQpU6wOQ0RERERERESkSdOYUiIiIiIiIiIi4nJKSomIiIiIiIiIiMspKSUiIiIiIiIiIi6nMaVcyDAMAPLz8y2ORERERBrS0Wf90We/mFQXEhERaRnqWxdSUsqFCgoKAGjTpo3FkYiIiIgrFBQUEBQUZHUYjYbqQiIiIi3Ln9WFbIa+wnMZh8NBSkoKAQEB2Gw2p147Pz+fNm3acPDgQQIDA5167aZGZVFDZVFDZVFDZVFDZVFDZVHDGWVhGAYFBQXExsZit2u0hKNUF3INlUUNlUUNlUVtKo8aKosaKosap1oW9a0LqaWUC9ntdlq3bt2g9wgMDGzx/3mOUlnUUFnUUFnUUFnUUFnUUFnUONWyUAup46ku5Foqixoqixoqi9pUHjVUFjVUFjVOpSzqUxfSV3ciIiIiIiIiIuJySkqJiIiIiIiIiIjLKSnVTHh5eTF16lS8vLysDsVyKosaKosaKosaKosaKosaKosaKoumSf9uNVQWNVQWNVQWtak8aqgsaqgsariqLDTQuYiIiIiIiIiIuJxaSomIiIiIiIiIiMspKSUiIiIiIiIiIi6npJSIiIiIiIiIiLicklLNxCuvvEJCQgLe3t4MHjyYVatWWR1Sg1u8eDHnnXcesbGx2Gw2Zs2aVeu4YRg8+uijxMTE4OPjw9ixY9m5c6c1wTagadOmMXDgQAICAoiMjOTCCy8kKSmp1jmlpaVMnjyZsLAw/P39mTBhAunp6RZF3HBee+01evXqRWBgIIGBgSQmJvLDDz9UH28p5VCXJ598EpvNxl133VW9r6WUx2OPPYbNZqu1dOnSpfp4SymHo5KTk7nqqqsICwvDx8eHnj17smbNmurjLeV3J0BCQsJxPxs2m43JkycDLe9no6lTXajl1oVA9aFjqT5Ut5ZcFwLVh35L9SFTY6gLKSnVDHz66afcc889TJ06lXXr1tG7d2/GjRtHRkaG1aE1qKKiInr37s0rr7xS5/Gnn36aF198kenTp7Ny5Ur8/PwYN24cpaWlLo60YS1atIjJkyezYsUK5s2bR0VFBWeeeSZFRUXV59x99918++23zJw5k0WLFpGSksLFF19sYdQNo3Xr1jz55JOsXbuWNWvWMHr0aC644AK2bNkCtJxy+K3Vq1fz+uuv06tXr1r7W1J5dO/endTU1OplyZIl1cdaUjnk5OQwdOhQPDw8+OGHH9i6dSvPPvssISEh1ee0lN+dYP7fOPbnYt68eQBceumlQMv62WjqVBdq2XUhUH3oWKoPHU91IZPqQybVh2o0irqQIU3eoEGDjMmTJ1e/rqqqMmJjY41p06ZZGJVrAcZXX31V/drhcBjR0dHGM888U70vNzfX8PLyMj7++GMLInSdjIwMAzAWLVpkGIb5uT08PIyZM2dWn7Nt2zYDMJYvX25VmC4TEhJivPXWWy22HAoKCoyOHTsa8+bNM0aOHGnceeedhmG0rJ+LqVOnGr17967zWEsqB8MwjAcffNAYNmzY7x5vyb87DcMw7rzzTqN9+/aGw+FocT8bTZ3qQqoL/ZbqQ7W15PqQ6kIm1YdqqD70+6yoC6mlVBNXXl7O2rVrGTt2bPU+u93O2LFjWb58uYWRWWvv3r2kpaXVKpegoCAGDx7c7MslLy8PgNDQUADWrl1LRUVFrbLo0qULcXFxzbosqqqq+OSTTygqKiIxMbHFlsPkyZM555xzan1uaHk/Fzt37iQ2NpZ27doxceJEDhw4ALS8cvjmm28YMGAAl156KZGRkfTt25c333yz+nhL/t1ZXl7OBx98wPXXX4/NZmtxPxtNmepCdWvJ/59B9aGjVB9SXehYqg+ZVB+qm1V1ISWlmrisrCyqqqqIioqqtT8qKoq0tDSLorLe0c/e0srF4XBw1113MXToUHr06AGYZeHp6UlwcHCtc5trWWzatAl/f3+8vLy49dZb+eqrr+jWrVuLKweATz75hHXr1jFt2rTjjrWk8hg8eDAzZsxgzpw5vPbaa+zdu5fhw4dTUFDQosoBYM+ePbz22mt07NiRuXPnctttt/HXv/6V9957D2i5vzsBZs2aRW5uLtdeey3Qsv6PNHWqC9WtJf9/Vn1I9aGjVBeqofpQDdWH6mZVXcjdaVcSEctNnjyZzZs31+of3tJ07tyZDRs2kJeXx+eff86kSZNYtGiR1WG53MGDB7nzzjuZN28e3t7eVodjqfHjx1dv9+rVi8GDBxMfH89nn32Gj4+PhZG5nsPhYMCAATzxxBMA9O3bl82bNzN9+nQmTZpkcXTWevvttxk/fjyxsbFWhyIip0j1IdWHQHWh31J9qIbqQ3Wzqi6kllJNXHh4OG5ubseNgJ+enk50dLRFUVnv6GdvSeUyZcoUvvvuO37++Wdat25dvT86Opry8nJyc3Nrnd9cy8LT05MOHTrQv39/pk2bRu/evfnvf//b4sph7dq1ZGRk0K9fP9zd3XF3d2fRokW8+OKLuLu7ExUV1aLK41jBwcF06tSJXbt2tbifi5iYGLp161ZrX9euXaub77fE350A+/fv56effuLGG2+s3tfSfjaaMtWF6tZS/z+rPmRSfUh1oT+j+pDqQ8eysi6kpFQT5+npSf/+/Zk/f371PofDwfz580lMTLQwMmu1bduW6OjoWuWSn5/PypUrm125GIbBlClT+Oqrr1iwYAFt27atdbx///54eHjUKoukpCQOHDjQ7MqiLg6Hg7KyshZXDmPGjGHTpk1s2LChehkwYAATJ06s3m5J5XGswsJCdu/eTUxMTIv7uRg6dOhxU6Tv2LGD+Ph4oGX97jzWu+++S2RkJOecc071vpb2s9GUqS5Ut5b2/1n1oT/WEutDqgv9MdWHVB86lqV1IacNmS6W+eSTTwwvLy9jxowZxtatW42bb77ZCA4ONtLS0qwOrUEVFBQY69evN9avX28AxnPPPWesX7/e2L9/v2EYhvHkk08awcHBxtdff21s3LjRuOCCC4y2bdsaJSUlFkfuXLfddpsRFBRkLFy40EhNTa1eiouLq8+59dZbjbi4OGPBggXGmjVrjMTERCMxMdHCqBvGQw89ZCxatMjYu3evsXHjRuOhhx4ybDab8eOPPxqG0XLK4fccO+OMYbSc8rj33nuNhQsXGnv37jWWLl1qjB071ggPDzcyMjIMw2g55WAYhrFq1SrD3d3d+Pe//23s3LnT+PDDDw1fX1/jgw8+qD6npfzuPKqqqsqIi4szHnzwweOOtaSfjaZOdaGWXRcyDNWHjqX60O9rqXUhw1B96FiqD9VmdV1ISalm4qWXXjLi4uIMT09PY9CgQcaKFSusDqnB/fzzzwZw3DJp0iTDMMypPP/xj38YUVFRhpeXlzFmzBgjKSnJ2qAbQF1lABjvvvtu9TklJSXG7bffboSEhBi+vr7GRRddZKSmploXdAO5/vrrjfj4eMPT09OIiIgwxowZU10BM4yWUw6/57cVsZZSHpdffrkRExNjeHp6Gq1atTIuv/xyY9euXdXHW0o5HPXtt98aPXr0MLy8vIwuXboYb7zxRq3jLeV351Fz5841gDo/Y0v72WjqVBdquXUhw1B96FiqD/2+lloXMgzVh35L9aEaVteFbIZhGM5rdyUiIiIiIiIiIvLnNKaUiIiIiIiIiIi4nJJSIiIiIiIiIiLickpKiYiIiIiIiIiIyykpJSIiIiIiIiIiLqeklIiIiIiIiIiIuJySUiIiIiIiIiIi4nJKSomIiIiIiIiIiMspKSUiIiIiIiIiIi6npJSISCNks9mYNWuW1WGIiIiIWEb1IZHmT0kpEZHfuPbaa7HZbMctZ511ltWhiYiIiLiE6kMi4gruVgcgItIYnXXWWbz77ru19nl5eVkUjYiIiIjrqT4kIg1NLaVEROrg5eVFdHR0rSUkJAQwm5K/9tprjB8/Hh8fH9q1a8fnn39e6/2bNm1i9OjR+Pj4EBYWxs0330xhYWGtc9555x26d++Ol5cXMTExTJkypdbxrKwsLrroInx9fenYsSPffPNNw35oERERkWOoPiQiDU1JKRGRk/CPf/yDCRMm8OuvvzJx4kSuuOIKtm3bBkBRURHjxo0jJCSE1atXM3PmTH766adalazXXnuNyZMnc/PNN7Np0ya++eYbOnToUOsejz/+OJdddhkbN27k7LPPZuLEiWRnZ7v0c4qIiIj8HtWHROSUGSIiUsukSZMMNzc3w8/Pr9by73//2zAMwwCMW2+9tdZ7Bg8ebNx2222GYRjGG2+8YYSEhBiFhYXVx2fPnm3Y7XYjLS3NMAzDiI2NNf7+97//bgyA8cgjj1S/LiwsNADjhx9+cNrnFBEREfk9qg+JiCtoTCkRkTqMGjWK1157rda+0NDQ6u3ExMRaxxITE9mwYQMA27Zto3fv3vj5+VUfHzp0KA6Hg6SkJGw2GykpKYwZM+YPY+jVq1f1tp+fH4GBgWRkZJzsRxIRERE5IaoPiUhDU1JKRKQOfn5+xzUfdxYfH596nefh4VHrtc1mw+FwNERIIiIiIsdRfUhEGprGlBIROQkrVqw47nXXrl0B6Nq1K7/++itFRUXVx5cuXYrdbqdz584EBASQkJDA/PnzXRqziIiIiDOpPiQip0otpURE6lBWVkZaWlqtfe7u7oSHhwMwc+ZMBgwYwLBhw/jwww9ZtWoVb7/9NgATJ05k6tSpTJo0iccee4zMzEzuuOMOrr76aqKiogB47LHHuPXWW4mMjGT8+PEUFBSwdOlS7rjjDtd+UBEREZHfofqQiDQ0JaVEROowZ84cYmJiau3r3Lkz27dvB8yZYD755BNuv/12YmJi+Pjjj+nWrRsAvr6+zJ07lzvvvJOBAwfi6+vLhAkTeO6556qvNWnSJEpLS3n++ee57777CA8P55JLLnHdBxQRERH5E6oPiUhDsxmGYVgdhIhIU2Kz2fjqq6+48MILrQ5FRERExBKqD4mIM2hMKRERERERERERcTklpURERERERERExOXUfU9ERERERERERFxOLaVERERERERERMTllJQSERERERERERGXU1JKRERERERERERcTkkpERERERERERFxOSWlRERERERERETE5ZSUEhERERERERERl1NSSkREREREREREXE5JKRERERERERERcTklpURERERERERExOX+H62e877vwblPAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1200x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "my_model.summary()\n",
    "\n",
    "fig, axs = plt.subplots(1, 2, figsize=(12, 6))\n",
    "\n",
    "# Training & validation loss values plot\n",
    "axs[0].plot(history.history['loss'], label='Training Loss')\n",
    "axs[0].plot(history.history['val_loss'], label='Validation Loss')\n",
    "axs[0].set_title('Model Loss')\n",
    "axs[0].set_xlabel('Epoch')\n",
    "axs[0].set_ylabel('Loss')\n",
    "axs[0].legend(['Train', 'Validation'], loc='upper right')\n",
    "\n",
    "# Training & validation mean absolute error values plot\n",
    "axs[1].plot(history.history['mean_absolute_error'], label='Training MAE')\n",
    "axs[1].plot(history.history['val_mean_absolute_error'], label='Validation MAE')\n",
    "axs[1].set_title('Mean Absolute Error')\n",
    "axs[1].set_xlabel('Epoch')\n",
    "axs[1].set_ylabel('MAE')\n",
    "axs[1].legend(['Train', 'Validation'], loc='upper right')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "80p35col880A"
   },
   "source": [
    "Actual vs. Predicted values:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 490
    },
    "id": "qSGRMBQJ8-cF",
    "outputId": "b5f0e6a5-2da4-410c-ddcc-ec5cee391942"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14/14 [==============================] - 0s 2ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred = my_model.predict(X_test)\n",
    "\n",
    "y_pred = y_pred.flatten()\n",
    "\n",
    "# create the scatter plot, zoomed in on a specific range\n",
    "plt.scatter(y_test, y_pred)\n",
    "plt.xlabel('Actual Values')\n",
    "plt.ylabel('Predicted Values')\n",
    "plt.title('Actual vs. Predicted Values (Zoomed In)')\n",
    "plt.xlim(0, 1000)\n",
    "plt.ylim(0, 1000)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "TXmJ1KaC9CXt"
   },
   "source": [
    "Error Distribution:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 472
    },
    "id": "ymRrqVvE9Jih",
    "outputId": "fd4d0fb1-c8af-4742-967a-3d08ebc62518"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred = y_pred.reshape(-1)\n",
    "\n",
    "error = y_test - y_pred\n",
    "plt.hist(error, bins=20)\n",
    "plt.xlabel('Prediction Error')\n",
    "plt.ylabel('Frequency')\n",
    "plt.title('Prediction Error Distribution')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7JsD9Nqs9QDu"
   },
   "source": [
    "K-fold validaton:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "Lgl8-MIK9Qtw",
    "outputId": "b09305f5-ed6a-486d-c7af-e899ff0fbc9e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "56/56 [==============================] - 1s 6ms/step - loss: 179854.7969 - mean_absolute_error: 250.1073 - val_loss: 182280.2188 - val_mean_absolute_error: 256.2039\n",
      "Epoch 2/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 173189.3281 - mean_absolute_error: 243.7929 - val_loss: 169255.3281 - val_mean_absolute_error: 245.0713\n",
      "Epoch 3/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 150438.7344 - mean_absolute_error: 224.8890 - val_loss: 133259.6562 - val_mean_absolute_error: 215.5978\n",
      "Epoch 4/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 104924.9922 - mean_absolute_error: 185.6131 - val_loss: 76899.3047 - val_mean_absolute_error: 161.2014\n",
      "Epoch 5/100\n",
      "56/56 [==============================] - 0s 2ms/step - loss: 52827.7578 - mean_absolute_error: 129.1051 - val_loss: 32670.1719 - val_mean_absolute_error: 101.3577\n",
      "Epoch 6/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 26026.5176 - mean_absolute_error: 95.9474 - val_loss: 19882.9395 - val_mean_absolute_error: 89.1749\n",
      "Epoch 7/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 19541.3691 - mean_absolute_error: 90.2707 - val_loss: 16770.1426 - val_mean_absolute_error: 85.5343\n",
      "Epoch 8/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 16963.9199 - mean_absolute_error: 85.7813 - val_loss: 14690.5361 - val_mean_absolute_error: 80.7391\n",
      "Epoch 9/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 15046.5781 - mean_absolute_error: 81.4367 - val_loss: 13009.4824 - val_mean_absolute_error: 76.5170\n",
      "Epoch 10/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 13462.7920 - mean_absolute_error: 77.4615 - val_loss: 11675.7129 - val_mean_absolute_error: 72.7673\n",
      "Epoch 11/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 12128.1982 - mean_absolute_error: 74.4002 - val_loss: 10530.0459 - val_mean_absolute_error: 69.7952\n",
      "Epoch 12/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 11012.1504 - mean_absolute_error: 71.5768 - val_loss: 9568.2051 - val_mean_absolute_error: 67.4167\n",
      "Epoch 13/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 10086.5674 - mean_absolute_error: 69.4289 - val_loss: 8814.6406 - val_mean_absolute_error: 65.5520\n",
      "Epoch 14/100\n",
      "56/56 [==============================] - 0s 2ms/step - loss: 9284.3936 - mean_absolute_error: 67.5499 - val_loss: 8199.9463 - val_mean_absolute_error: 64.0719\n",
      "Epoch 15/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 8617.6611 - mean_absolute_error: 65.7762 - val_loss: 7683.9404 - val_mean_absolute_error: 62.8084\n",
      "Epoch 16/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 8056.4097 - mean_absolute_error: 64.2773 - val_loss: 7246.6636 - val_mean_absolute_error: 61.5436\n",
      "Epoch 17/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 7579.0830 - mean_absolute_error: 62.6965 - val_loss: 6898.6128 - val_mean_absolute_error: 60.4863\n",
      "Epoch 18/100\n",
      "56/56 [==============================] - 1s 25ms/step - loss: 7175.2300 - mean_absolute_error: 61.2670 - val_loss: 6616.2734 - val_mean_absolute_error: 59.4149\n",
      "Epoch 19/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 6849.5562 - mean_absolute_error: 60.0090 - val_loss: 6341.8950 - val_mean_absolute_error: 58.4291\n",
      "Epoch 20/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 6530.5142 - mean_absolute_error: 58.6125 - val_loss: 6126.3320 - val_mean_absolute_error: 57.3357\n",
      "Epoch 21/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 6269.1792 - mean_absolute_error: 57.3789 - val_loss: 5909.1089 - val_mean_absolute_error: 56.2230\n",
      "Epoch 22/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 6011.8423 - mean_absolute_error: 56.2316 - val_loss: 5695.2881 - val_mean_absolute_error: 55.1948\n",
      "Epoch 23/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 5779.8608 - mean_absolute_error: 55.0671 - val_loss: 5487.4609 - val_mean_absolute_error: 54.0472\n",
      "Epoch 24/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 5541.4482 - mean_absolute_error: 53.7322 - val_loss: 5278.3247 - val_mean_absolute_error: 52.7961\n",
      "Epoch 25/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 5299.9326 - mean_absolute_error: 52.2779 - val_loss: 5053.9126 - val_mean_absolute_error: 51.2430\n",
      "Epoch 26/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 5028.5386 - mean_absolute_error: 50.6523 - val_loss: 4815.3330 - val_mean_absolute_error: 49.7763\n",
      "Epoch 27/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4789.1987 - mean_absolute_error: 48.9709 - val_loss: 4584.5127 - val_mean_absolute_error: 48.1508\n",
      "Epoch 28/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4533.3496 - mean_absolute_error: 47.1737 - val_loss: 4341.7588 - val_mean_absolute_error: 46.4145\n",
      "Epoch 29/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4278.8633 - mean_absolute_error: 45.8157 - val_loss: 4093.5203 - val_mean_absolute_error: 44.7963\n",
      "Epoch 30/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4034.1140 - mean_absolute_error: 44.0124 - val_loss: 3855.7122 - val_mean_absolute_error: 43.0027\n",
      "Epoch 31/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3775.5435 - mean_absolute_error: 42.6585 - val_loss: 3593.5442 - val_mean_absolute_error: 41.2533\n",
      "Epoch 32/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3512.0278 - mean_absolute_error: 40.7503 - val_loss: 3346.7415 - val_mean_absolute_error: 39.6431\n",
      "Epoch 33/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3258.3870 - mean_absolute_error: 39.1460 - val_loss: 3107.0723 - val_mean_absolute_error: 37.8882\n",
      "Epoch 34/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3002.4683 - mean_absolute_error: 37.5870 - val_loss: 2858.5391 - val_mean_absolute_error: 36.3950\n",
      "Epoch 35/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2761.7019 - mean_absolute_error: 36.0940 - val_loss: 2627.5469 - val_mean_absolute_error: 34.8442\n",
      "Epoch 36/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2522.3083 - mean_absolute_error: 34.3385 - val_loss: 2407.6460 - val_mean_absolute_error: 33.2010\n",
      "Epoch 37/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2296.1675 - mean_absolute_error: 32.9073 - val_loss: 2182.5562 - val_mean_absolute_error: 31.3954\n",
      "Epoch 38/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2079.8845 - mean_absolute_error: 31.2382 - val_loss: 1987.6750 - val_mean_absolute_error: 30.3367\n",
      "Epoch 39/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1882.0375 - mean_absolute_error: 29.7225 - val_loss: 1784.2980 - val_mean_absolute_error: 28.3229\n",
      "Epoch 40/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1679.4210 - mean_absolute_error: 27.9708 - val_loss: 1627.6964 - val_mean_absolute_error: 26.8683\n",
      "Epoch 41/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1510.6135 - mean_absolute_error: 26.7040 - val_loss: 1458.8514 - val_mean_absolute_error: 25.2038\n",
      "Epoch 42/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1354.3795 - mean_absolute_error: 24.9524 - val_loss: 1335.5499 - val_mean_absolute_error: 23.7130\n",
      "Epoch 43/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1209.4675 - mean_absolute_error: 23.5924 - val_loss: 1178.6056 - val_mean_absolute_error: 22.7282\n",
      "Epoch 44/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 1078.3920 - mean_absolute_error: 22.0340 - val_loss: 1093.5254 - val_mean_absolute_error: 21.2236\n",
      "Epoch 45/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 977.9218 - mean_absolute_error: 20.8514 - val_loss: 994.4776 - val_mean_absolute_error: 20.0793\n",
      "Epoch 46/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 869.7731 - mean_absolute_error: 19.5793 - val_loss: 879.9291 - val_mean_absolute_error: 18.6823\n",
      "Epoch 47/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 783.0175 - mean_absolute_error: 18.3024 - val_loss: 820.8615 - val_mean_absolute_error: 17.7904\n",
      "Epoch 48/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 715.1351 - mean_absolute_error: 17.3669 - val_loss: 738.3776 - val_mean_absolute_error: 16.7440\n",
      "Epoch 49/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 649.2260 - mean_absolute_error: 16.5107 - val_loss: 682.3498 - val_mean_absolute_error: 16.0359\n",
      "Epoch 50/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 589.8849 - mean_absolute_error: 15.7357 - val_loss: 633.2068 - val_mean_absolute_error: 15.4924\n",
      "Epoch 51/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 548.6488 - mean_absolute_error: 15.0072 - val_loss: 588.6152 - val_mean_absolute_error: 14.7987\n",
      "Epoch 52/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 508.0029 - mean_absolute_error: 14.3408 - val_loss: 570.1976 - val_mean_absolute_error: 14.3400\n",
      "Epoch 53/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 479.6919 - mean_absolute_error: 13.9845 - val_loss: 527.3176 - val_mean_absolute_error: 13.9974\n",
      "Epoch 54/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 447.2467 - mean_absolute_error: 13.5841 - val_loss: 518.0167 - val_mean_absolute_error: 13.5747\n",
      "Epoch 55/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 423.1949 - mean_absolute_error: 13.2163 - val_loss: 493.4781 - val_mean_absolute_error: 13.2108\n",
      "Epoch 56/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 405.4106 - mean_absolute_error: 13.0052 - val_loss: 466.9809 - val_mean_absolute_error: 13.0700\n",
      "Epoch 57/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 389.4532 - mean_absolute_error: 12.6120 - val_loss: 471.7796 - val_mean_absolute_error: 12.7662\n",
      "Epoch 58/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 368.4953 - mean_absolute_error: 12.3817 - val_loss: 441.9236 - val_mean_absolute_error: 12.6578\n",
      "Epoch 59/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 358.6219 - mean_absolute_error: 12.2982 - val_loss: 434.1286 - val_mean_absolute_error: 12.4038\n",
      "Epoch 60/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 342.4126 - mean_absolute_error: 11.9144 - val_loss: 431.0150 - val_mean_absolute_error: 12.5636\n",
      "Epoch 61/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 331.5585 - mean_absolute_error: 11.9812 - val_loss: 412.0107 - val_mean_absolute_error: 12.0466\n",
      "Epoch 62/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 323.9533 - mean_absolute_error: 11.7972 - val_loss: 409.1591 - val_mean_absolute_error: 11.9991\n",
      "Epoch 63/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 312.9272 - mean_absolute_error: 11.5263 - val_loss: 392.3210 - val_mean_absolute_error: 12.0195\n",
      "Epoch 64/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 305.5064 - mean_absolute_error: 11.4600 - val_loss: 385.4023 - val_mean_absolute_error: 11.9035\n",
      "Epoch 65/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 293.8284 - mean_absolute_error: 11.2103 - val_loss: 375.8722 - val_mean_absolute_error: 11.7164\n",
      "Epoch 66/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 289.9368 - mean_absolute_error: 11.0214 - val_loss: 373.2187 - val_mean_absolute_error: 11.4876\n",
      "Epoch 67/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 278.3600 - mean_absolute_error: 10.9387 - val_loss: 369.9819 - val_mean_absolute_error: 11.4499\n",
      "Epoch 68/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 271.6137 - mean_absolute_error: 10.8048 - val_loss: 374.0768 - val_mean_absolute_error: 11.3819\n",
      "Epoch 69/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 261.9838 - mean_absolute_error: 10.6226 - val_loss: 348.7530 - val_mean_absolute_error: 11.1239\n",
      "Epoch 70/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 252.4331 - mean_absolute_error: 10.4326 - val_loss: 341.0979 - val_mean_absolute_error: 11.1213\n",
      "Epoch 71/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 249.1842 - mean_absolute_error: 10.3937 - val_loss: 362.0898 - val_mean_absolute_error: 11.0682\n",
      "Epoch 72/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 240.2583 - mean_absolute_error: 10.0825 - val_loss: 332.7451 - val_mean_absolute_error: 10.7908\n",
      "Epoch 73/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 235.2855 - mean_absolute_error: 10.0304 - val_loss: 331.1141 - val_mean_absolute_error: 10.7098\n",
      "Epoch 74/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 228.8667 - mean_absolute_error: 9.8728 - val_loss: 337.7690 - val_mean_absolute_error: 10.6345\n",
      "Epoch 75/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 220.8679 - mean_absolute_error: 9.7079 - val_loss: 340.2632 - val_mean_absolute_error: 10.7489\n",
      "Epoch 76/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 216.1052 - mean_absolute_error: 9.5117 - val_loss: 305.4946 - val_mean_absolute_error: 10.0785\n",
      "Epoch 77/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 208.3311 - mean_absolute_error: 9.3923 - val_loss: 301.4936 - val_mean_absolute_error: 10.0393\n",
      "Epoch 78/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 206.6349 - mean_absolute_error: 9.3367 - val_loss: 296.7155 - val_mean_absolute_error: 9.8676\n",
      "Epoch 79/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 201.9208 - mean_absolute_error: 9.2186 - val_loss: 283.6659 - val_mean_absolute_error: 9.7651\n",
      "Epoch 80/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 195.5033 - mean_absolute_error: 9.1732 - val_loss: 317.3606 - val_mean_absolute_error: 10.3307\n",
      "Epoch 81/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 192.0163 - mean_absolute_error: 8.9788 - val_loss: 276.5088 - val_mean_absolute_error: 9.4406\n",
      "Epoch 82/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 185.0678 - mean_absolute_error: 8.7866 - val_loss: 275.2545 - val_mean_absolute_error: 9.4999\n",
      "Epoch 83/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 179.4891 - mean_absolute_error: 8.6343 - val_loss: 272.1310 - val_mean_absolute_error: 9.2796\n",
      "Epoch 84/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 174.3565 - mean_absolute_error: 8.4700 - val_loss: 267.2491 - val_mean_absolute_error: 9.1567\n",
      "Epoch 85/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 170.5548 - mean_absolute_error: 8.3519 - val_loss: 250.3069 - val_mean_absolute_error: 8.7005\n",
      "Epoch 86/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 166.6552 - mean_absolute_error: 8.2003 - val_loss: 251.2041 - val_mean_absolute_error: 8.8919\n",
      "Epoch 87/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 165.3433 - mean_absolute_error: 8.2398 - val_loss: 253.5358 - val_mean_absolute_error: 8.7325\n",
      "Epoch 88/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 156.9345 - mean_absolute_error: 7.8973 - val_loss: 237.1606 - val_mean_absolute_error: 8.4388\n",
      "Epoch 89/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 153.7867 - mean_absolute_error: 7.8300 - val_loss: 231.8504 - val_mean_absolute_error: 8.3514\n",
      "Epoch 90/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 149.5090 - mean_absolute_error: 7.8397 - val_loss: 237.7866 - val_mean_absolute_error: 8.3076\n",
      "Epoch 91/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 147.8088 - mean_absolute_error: 7.6271 - val_loss: 223.3075 - val_mean_absolute_error: 8.0732\n",
      "Epoch 92/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 138.7844 - mean_absolute_error: 7.3912 - val_loss: 222.4781 - val_mean_absolute_error: 7.9693\n",
      "Epoch 93/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 136.4609 - mean_absolute_error: 7.3233 - val_loss: 217.3323 - val_mean_absolute_error: 8.2009\n",
      "Epoch 94/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 131.3686 - mean_absolute_error: 7.1333 - val_loss: 208.5903 - val_mean_absolute_error: 7.8391\n",
      "Epoch 95/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 128.4763 - mean_absolute_error: 7.0218 - val_loss: 205.8861 - val_mean_absolute_error: 7.8040\n",
      "Epoch 96/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 124.6068 - mean_absolute_error: 7.0060 - val_loss: 191.2131 - val_mean_absolute_error: 7.3432\n",
      "Epoch 97/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 122.1901 - mean_absolute_error: 6.8814 - val_loss: 194.6093 - val_mean_absolute_error: 7.3011\n",
      "Epoch 98/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 116.0420 - mean_absolute_error: 6.6507 - val_loss: 183.6854 - val_mean_absolute_error: 7.0031\n",
      "Epoch 99/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 112.7447 - mean_absolute_error: 6.5776 - val_loss: 180.9276 - val_mean_absolute_error: 7.1774\n",
      "Epoch 100/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 109.2694 - mean_absolute_error: 6.4618 - val_loss: 173.5984 - val_mean_absolute_error: 6.7712\n",
      "14/14 [==============================] - 0s 2ms/step - loss: 173.5984 - mean_absolute_error: 6.7712\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "56/56 [==============================] - 1s 5ms/step - loss: 182415.7031 - mean_absolute_error: 252.0975 - val_loss: 176450.0000 - val_mean_absolute_error: 251.5408\n",
      "Epoch 2/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 177772.2031 - mean_absolute_error: 247.5249 - val_loss: 167145.5156 - val_mean_absolute_error: 243.4084\n",
      "Epoch 3/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 159712.9688 - mean_absolute_error: 232.3566 - val_loss: 138058.4062 - val_mean_absolute_error: 219.9655\n",
      "Epoch 4/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 118986.2422 - mean_absolute_error: 197.8825 - val_loss: 86357.4453 - val_mean_absolute_error: 173.9252\n",
      "Epoch 5/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 66394.8359 - mean_absolute_error: 144.6027 - val_loss: 40193.9492 - val_mean_absolute_error: 116.9484\n",
      "Epoch 6/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 32143.5293 - mean_absolute_error: 102.0655 - val_loss: 21959.6523 - val_mean_absolute_error: 92.1619\n",
      "Epoch 7/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 21800.4551 - mean_absolute_error: 93.6050 - val_loss: 18367.7559 - val_mean_absolute_error: 87.8746\n",
      "Epoch 8/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 18289.7246 - mean_absolute_error: 88.6318 - val_loss: 15972.1182 - val_mean_absolute_error: 84.5478\n",
      "Epoch 9/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 15815.1689 - mean_absolute_error: 84.8646 - val_loss: 13878.9395 - val_mean_absolute_error: 81.3794\n",
      "Epoch 10/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 13709.9697 - mean_absolute_error: 81.0610 - val_loss: 12209.2910 - val_mean_absolute_error: 78.1207\n",
      "Epoch 11/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 12072.0391 - mean_absolute_error: 77.5832 - val_loss: 10890.2842 - val_mean_absolute_error: 75.6845\n",
      "Epoch 12/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 10693.5342 - mean_absolute_error: 74.8186 - val_loss: 9707.2510 - val_mean_absolute_error: 72.1703\n",
      "Epoch 13/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 9628.6191 - mean_absolute_error: 71.9447 - val_loss: 8791.6123 - val_mean_absolute_error: 69.6073\n",
      "Epoch 14/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 8722.7305 - mean_absolute_error: 69.6439 - val_loss: 8060.3242 - val_mean_absolute_error: 67.2844\n",
      "Epoch 15/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 8007.0156 - mean_absolute_error: 67.2066 - val_loss: 7356.7607 - val_mean_absolute_error: 64.3865\n",
      "Epoch 16/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 7427.2734 - mean_absolute_error: 64.9152 - val_loss: 6824.7954 - val_mean_absolute_error: 62.4462\n",
      "Epoch 17/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6921.7026 - mean_absolute_error: 62.9466 - val_loss: 6391.8911 - val_mean_absolute_error: 60.5315\n",
      "Epoch 18/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6486.2314 - mean_absolute_error: 60.9084 - val_loss: 6014.9131 - val_mean_absolute_error: 58.7290\n",
      "Epoch 19/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6101.6758 - mean_absolute_error: 59.0065 - val_loss: 5580.2266 - val_mean_absolute_error: 56.4475\n",
      "Epoch 20/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 5743.3564 - mean_absolute_error: 56.9968 - val_loss: 5233.4609 - val_mean_absolute_error: 54.4925\n",
      "Epoch 21/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 5408.6631 - mean_absolute_error: 55.1651 - val_loss: 4868.3359 - val_mean_absolute_error: 52.2552\n",
      "Epoch 22/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 5065.5518 - mean_absolute_error: 53.0091 - val_loss: 4561.4502 - val_mean_absolute_error: 50.2865\n",
      "Epoch 23/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4741.6802 - mean_absolute_error: 50.4470 - val_loss: 4216.4585 - val_mean_absolute_error: 47.6213\n",
      "Epoch 24/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4427.3916 - mean_absolute_error: 48.2165 - val_loss: 3931.5566 - val_mean_absolute_error: 45.6738\n",
      "Epoch 25/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 4129.2803 - mean_absolute_error: 45.9167 - val_loss: 3611.3711 - val_mean_absolute_error: 43.0952\n",
      "Epoch 26/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3832.4749 - mean_absolute_error: 43.8244 - val_loss: 3343.7319 - val_mean_absolute_error: 41.3090\n",
      "Epoch 27/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3554.0725 - mean_absolute_error: 41.8790 - val_loss: 3122.4812 - val_mean_absolute_error: 39.9202\n",
      "Epoch 28/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3304.2908 - mean_absolute_error: 40.2630 - val_loss: 2861.3618 - val_mean_absolute_error: 37.9950\n",
      "Epoch 29/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3044.6919 - mean_absolute_error: 38.5230 - val_loss: 2617.3469 - val_mean_absolute_error: 36.2499\n",
      "Epoch 30/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2803.0757 - mean_absolute_error: 36.8179 - val_loss: 2387.2524 - val_mean_absolute_error: 34.5279\n",
      "Epoch 31/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2579.4146 - mean_absolute_error: 35.3472 - val_loss: 2171.8401 - val_mean_absolute_error: 32.7445\n",
      "Epoch 32/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2369.2732 - mean_absolute_error: 33.6874 - val_loss: 1975.4833 - val_mean_absolute_error: 31.2788\n",
      "Epoch 33/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 2166.0073 - mean_absolute_error: 32.2014 - val_loss: 1785.3763 - val_mean_absolute_error: 29.8692\n",
      "Epoch 34/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1974.8641 - mean_absolute_error: 30.8321 - val_loss: 1633.2858 - val_mean_absolute_error: 28.0817\n",
      "Epoch 35/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1803.7565 - mean_absolute_error: 28.9556 - val_loss: 1465.8226 - val_mean_absolute_error: 26.7003\n",
      "Epoch 36/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1655.0546 - mean_absolute_error: 27.8598 - val_loss: 1331.6259 - val_mean_absolute_error: 24.6443\n",
      "Epoch 37/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1503.8182 - mean_absolute_error: 25.9825 - val_loss: 1204.8281 - val_mean_absolute_error: 23.8397\n",
      "Epoch 38/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1376.2367 - mean_absolute_error: 24.7915 - val_loss: 1091.3076 - val_mean_absolute_error: 22.1089\n",
      "Epoch 39/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1245.0281 - mean_absolute_error: 23.3337 - val_loss: 986.0460 - val_mean_absolute_error: 21.3258\n",
      "Epoch 40/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1143.0017 - mean_absolute_error: 22.2782 - val_loss: 898.4019 - val_mean_absolute_error: 19.6996\n",
      "Epoch 41/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1042.7266 - mean_absolute_error: 20.9477 - val_loss: 823.1985 - val_mean_absolute_error: 18.5906\n",
      "Epoch 42/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 953.6609 - mean_absolute_error: 19.7473 - val_loss: 747.3247 - val_mean_absolute_error: 17.8471\n",
      "Epoch 43/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 877.8056 - mean_absolute_error: 18.8431 - val_loss: 689.9333 - val_mean_absolute_error: 16.8579\n",
      "Epoch 44/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 806.0093 - mean_absolute_error: 17.8368 - val_loss: 626.8651 - val_mean_absolute_error: 15.8679\n",
      "Epoch 45/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 745.3429 - mean_absolute_error: 17.0135 - val_loss: 612.5617 - val_mean_absolute_error: 16.6891\n",
      "Epoch 46/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 704.9840 - mean_absolute_error: 16.6389 - val_loss: 537.0276 - val_mean_absolute_error: 13.6256\n",
      "Epoch 47/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 646.9603 - mean_absolute_error: 15.6412 - val_loss: 502.8596 - val_mean_absolute_error: 13.7191\n",
      "Epoch 48/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 604.9541 - mean_absolute_error: 15.0732 - val_loss: 467.4522 - val_mean_absolute_error: 13.6412\n",
      "Epoch 49/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 567.0950 - mean_absolute_error: 14.5965 - val_loss: 443.1902 - val_mean_absolute_error: 13.2621\n",
      "Epoch 50/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 528.8954 - mean_absolute_error: 14.0316 - val_loss: 412.1523 - val_mean_absolute_error: 12.4477\n",
      "Epoch 51/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 508.2162 - mean_absolute_error: 13.8500 - val_loss: 396.9131 - val_mean_absolute_error: 12.7548\n",
      "Epoch 52/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 477.2603 - mean_absolute_error: 13.3643 - val_loss: 371.8309 - val_mean_absolute_error: 11.7485\n",
      "Epoch 53/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 452.0459 - mean_absolute_error: 12.9599 - val_loss: 355.1805 - val_mean_absolute_error: 11.8815\n",
      "Epoch 54/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 430.7288 - mean_absolute_error: 12.6080 - val_loss: 334.0273 - val_mean_absolute_error: 10.8879\n",
      "Epoch 55/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 412.0762 - mean_absolute_error: 12.5300 - val_loss: 325.1805 - val_mean_absolute_error: 10.7219\n",
      "Epoch 56/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 392.0769 - mean_absolute_error: 12.1350 - val_loss: 333.7647 - val_mean_absolute_error: 11.8319\n",
      "Epoch 57/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 375.5246 - mean_absolute_error: 11.7928 - val_loss: 328.4160 - val_mean_absolute_error: 12.2198\n",
      "Epoch 58/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 361.8276 - mean_absolute_error: 11.6847 - val_loss: 283.9525 - val_mean_absolute_error: 10.0883\n",
      "Epoch 59/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 342.6837 - mean_absolute_error: 11.1808 - val_loss: 272.0486 - val_mean_absolute_error: 9.9768\n",
      "Epoch 60/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 333.3018 - mean_absolute_error: 11.2579 - val_loss: 260.4797 - val_mean_absolute_error: 9.7101\n",
      "Epoch 61/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 316.5423 - mean_absolute_error: 10.9292 - val_loss: 269.7543 - val_mean_absolute_error: 10.3673\n",
      "Epoch 62/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 308.0746 - mean_absolute_error: 10.7605 - val_loss: 241.6054 - val_mean_absolute_error: 9.2857\n",
      "Epoch 63/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 295.2360 - mean_absolute_error: 10.4711 - val_loss: 233.4074 - val_mean_absolute_error: 9.2281\n",
      "Epoch 64/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 284.6678 - mean_absolute_error: 10.3066 - val_loss: 226.0338 - val_mean_absolute_error: 9.0093\n",
      "Epoch 65/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 275.7874 - mean_absolute_error: 10.2367 - val_loss: 221.9845 - val_mean_absolute_error: 9.0203\n",
      "Epoch 66/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 266.1212 - mean_absolute_error: 10.0436 - val_loss: 215.8632 - val_mean_absolute_error: 8.8004\n",
      "Epoch 67/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 260.9509 - mean_absolute_error: 9.8102 - val_loss: 205.9485 - val_mean_absolute_error: 8.5366\n",
      "Epoch 68/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 247.3535 - mean_absolute_error: 9.6320 - val_loss: 197.2362 - val_mean_absolute_error: 8.4069\n",
      "Epoch 69/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 244.3515 - mean_absolute_error: 9.4869 - val_loss: 191.4532 - val_mean_absolute_error: 8.3316\n",
      "Epoch 70/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 236.8304 - mean_absolute_error: 9.3025 - val_loss: 183.5592 - val_mean_absolute_error: 8.1257\n",
      "Epoch 71/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 229.4063 - mean_absolute_error: 9.3156 - val_loss: 180.5863 - val_mean_absolute_error: 8.1078\n",
      "Epoch 72/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 223.2987 - mean_absolute_error: 9.0854 - val_loss: 172.0536 - val_mean_absolute_error: 7.8169\n",
      "Epoch 73/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 215.4973 - mean_absolute_error: 8.9220 - val_loss: 190.9356 - val_mean_absolute_error: 9.2932\n",
      "Epoch 74/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 206.0441 - mean_absolute_error: 8.8465 - val_loss: 169.2991 - val_mean_absolute_error: 7.7595\n",
      "Epoch 75/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 198.9868 - mean_absolute_error: 8.4535 - val_loss: 161.4303 - val_mean_absolute_error: 7.9504\n",
      "Epoch 76/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 198.4270 - mean_absolute_error: 8.5487 - val_loss: 152.0440 - val_mean_absolute_error: 7.5902\n",
      "Epoch 77/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 186.3989 - mean_absolute_error: 8.3258 - val_loss: 145.0091 - val_mean_absolute_error: 7.2255\n",
      "Epoch 78/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 182.9363 - mean_absolute_error: 8.2621 - val_loss: 140.0775 - val_mean_absolute_error: 7.1977\n",
      "Epoch 79/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 177.3494 - mean_absolute_error: 8.0862 - val_loss: 145.7888 - val_mean_absolute_error: 7.2138\n",
      "Epoch 80/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 170.1027 - mean_absolute_error: 7.8755 - val_loss: 147.5850 - val_mean_absolute_error: 7.9568\n",
      "Epoch 81/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 161.8198 - mean_absolute_error: 7.6427 - val_loss: 123.4269 - val_mean_absolute_error: 6.8082\n",
      "Epoch 82/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 154.0578 - mean_absolute_error: 7.4652 - val_loss: 117.6982 - val_mean_absolute_error: 6.6205\n",
      "Epoch 83/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 147.1882 - mean_absolute_error: 7.3273 - val_loss: 112.7078 - val_mean_absolute_error: 6.3311\n",
      "Epoch 84/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 141.1245 - mean_absolute_error: 7.0975 - val_loss: 107.4913 - val_mean_absolute_error: 6.2594\n",
      "Epoch 85/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 138.0384 - mean_absolute_error: 7.0037 - val_loss: 102.4701 - val_mean_absolute_error: 6.1219\n",
      "Epoch 86/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 131.5461 - mean_absolute_error: 7.0016 - val_loss: 98.2115 - val_mean_absolute_error: 5.9699\n",
      "Epoch 87/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 126.6986 - mean_absolute_error: 6.7128 - val_loss: 94.6847 - val_mean_absolute_error: 5.9419\n",
      "Epoch 88/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 122.6636 - mean_absolute_error: 6.6504 - val_loss: 105.7786 - val_mean_absolute_error: 6.8521\n",
      "Epoch 89/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 120.0377 - mean_absolute_error: 6.5496 - val_loss: 86.4836 - val_mean_absolute_error: 5.6333\n",
      "Epoch 90/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 112.1873 - mean_absolute_error: 6.2696 - val_loss: 84.4522 - val_mean_absolute_error: 5.7426\n",
      "Epoch 91/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 105.5334 - mean_absolute_error: 6.0694 - val_loss: 80.8727 - val_mean_absolute_error: 5.3458\n",
      "Epoch 92/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 101.6282 - mean_absolute_error: 5.9101 - val_loss: 77.2153 - val_mean_absolute_error: 5.4866\n",
      "Epoch 93/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 98.7530 - mean_absolute_error: 5.8519 - val_loss: 72.2698 - val_mean_absolute_error: 5.2430\n",
      "Epoch 94/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 94.0138 - mean_absolute_error: 5.7251 - val_loss: 71.5250 - val_mean_absolute_error: 5.0542\n",
      "Epoch 95/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 95.2314 - mean_absolute_error: 5.8240 - val_loss: 70.5202 - val_mean_absolute_error: 5.4784\n",
      "Epoch 96/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 87.7077 - mean_absolute_error: 5.5224 - val_loss: 62.7357 - val_mean_absolute_error: 4.8518\n",
      "Epoch 97/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 82.7812 - mean_absolute_error: 5.2922 - val_loss: 59.0383 - val_mean_absolute_error: 4.5935\n",
      "Epoch 98/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 80.8491 - mean_absolute_error: 5.2660 - val_loss: 61.9407 - val_mean_absolute_error: 4.7638\n",
      "Epoch 99/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 75.1213 - mean_absolute_error: 5.0497 - val_loss: 54.3835 - val_mean_absolute_error: 4.4175\n",
      "Epoch 100/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 72.8250 - mean_absolute_error: 4.9388 - val_loss: 62.3945 - val_mean_absolute_error: 4.8119\n",
      "14/14 [==============================] - 0s 3ms/step - loss: 62.3945 - mean_absolute_error: 4.8119\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "56/56 [==============================] - 1s 5ms/step - loss: 180500.7969 - mean_absolute_error: 251.7164 - val_loss: 183736.8281 - val_mean_absolute_error: 254.4966\n",
      "Epoch 2/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 176285.2969 - mean_absolute_error: 248.3490 - val_loss: 175599.4062 - val_mean_absolute_error: 247.5354\n",
      "Epoch 3/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 161776.5156 - mean_absolute_error: 236.7141 - val_loss: 152824.5781 - val_mean_absolute_error: 228.5101\n",
      "Epoch 4/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 129482.5469 - mean_absolute_error: 210.0585 - val_loss: 110878.7109 - val_mean_absolute_error: 191.5112\n",
      "Epoch 5/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 83003.9375 - mean_absolute_error: 164.4546 - val_loss: 63590.9961 - val_mean_absolute_error: 140.1716\n",
      "Epoch 6/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 43374.1523 - mean_absolute_error: 113.9491 - val_loss: 32334.0098 - val_mean_absolute_error: 100.3158\n",
      "Epoch 7/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 24093.7598 - mean_absolute_error: 88.8271 - val_loss: 22056.0293 - val_mean_absolute_error: 91.9525\n",
      "Epoch 8/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 18290.6484 - mean_absolute_error: 83.4979 - val_loss: 18512.6191 - val_mean_absolute_error: 89.7483\n",
      "Epoch 9/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 15688.8662 - mean_absolute_error: 80.5782 - val_loss: 16213.2373 - val_mean_absolute_error: 85.4499\n",
      "Epoch 10/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 13706.8926 - mean_absolute_error: 76.6362 - val_loss: 14421.0049 - val_mean_absolute_error: 81.3937\n",
      "Epoch 11/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 12176.3174 - mean_absolute_error: 72.7803 - val_loss: 12963.1348 - val_mean_absolute_error: 77.8666\n",
      "Epoch 12/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 10947.6973 - mean_absolute_error: 70.2867 - val_loss: 11686.0996 - val_mean_absolute_error: 74.7007\n",
      "Epoch 13/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 9853.6787 - mean_absolute_error: 67.6505 - val_loss: 10680.8623 - val_mean_absolute_error: 72.4288\n",
      "Epoch 14/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 9035.4873 - mean_absolute_error: 65.9217 - val_loss: 9755.6787 - val_mean_absolute_error: 69.6740\n",
      "Epoch 15/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 8309.4941 - mean_absolute_error: 63.8596 - val_loss: 9041.5361 - val_mean_absolute_error: 67.4919\n",
      "Epoch 16/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 7714.2368 - mean_absolute_error: 61.9692 - val_loss: 8425.5303 - val_mean_absolute_error: 65.2643\n",
      "Epoch 17/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 7220.5669 - mean_absolute_error: 60.5150 - val_loss: 7889.5708 - val_mean_absolute_error: 63.4184\n",
      "Epoch 18/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6809.1650 - mean_absolute_error: 59.1670 - val_loss: 7454.6851 - val_mean_absolute_error: 61.7292\n",
      "Epoch 19/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6447.4604 - mean_absolute_error: 57.7627 - val_loss: 7078.7080 - val_mean_absolute_error: 60.0844\n",
      "Epoch 20/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6118.4170 - mean_absolute_error: 56.5497 - val_loss: 6730.6079 - val_mean_absolute_error: 58.5054\n",
      "Epoch 21/100\n",
      "56/56 [==============================] - 0s 2ms/step - loss: 5854.9194 - mean_absolute_error: 55.3368 - val_loss: 6394.9937 - val_mean_absolute_error: 56.9556\n",
      "Epoch 22/100\n",
      "56/56 [==============================] - 0s 2ms/step - loss: 5583.9541 - mean_absolute_error: 53.9827 - val_loss: 6123.2231 - val_mean_absolute_error: 55.4941\n",
      "Epoch 23/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 5336.5742 - mean_absolute_error: 52.6061 - val_loss: 5846.8423 - val_mean_absolute_error: 54.0498\n",
      "Epoch 24/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 5103.0601 - mean_absolute_error: 51.6356 - val_loss: 5581.4429 - val_mean_absolute_error: 52.8277\n",
      "Epoch 25/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4852.4917 - mean_absolute_error: 50.0741 - val_loss: 5354.8877 - val_mean_absolute_error: 51.6424\n",
      "Epoch 26/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 4628.7886 - mean_absolute_error: 49.1222 - val_loss: 5089.2319 - val_mean_absolute_error: 50.1389\n",
      "Epoch 27/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4385.6372 - mean_absolute_error: 47.6061 - val_loss: 4898.3374 - val_mean_absolute_error: 49.5109\n",
      "Epoch 28/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4170.7412 - mean_absolute_error: 46.5958 - val_loss: 4610.4717 - val_mean_absolute_error: 47.6108\n",
      "Epoch 29/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3950.4517 - mean_absolute_error: 45.2714 - val_loss: 4396.8921 - val_mean_absolute_error: 46.7525\n",
      "Epoch 30/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3717.1775 - mean_absolute_error: 43.9425 - val_loss: 4145.2119 - val_mean_absolute_error: 45.4274\n",
      "Epoch 31/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3503.0239 - mean_absolute_error: 42.6832 - val_loss: 3907.6948 - val_mean_absolute_error: 44.0376\n",
      "Epoch 32/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 3277.0737 - mean_absolute_error: 41.2561 - val_loss: 3707.6106 - val_mean_absolute_error: 42.9739\n",
      "Epoch 33/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3065.7632 - mean_absolute_error: 40.1855 - val_loss: 3463.0034 - val_mean_absolute_error: 41.6907\n",
      "Epoch 34/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2854.4785 - mean_absolute_error: 38.6869 - val_loss: 3261.7539 - val_mean_absolute_error: 40.3688\n",
      "Epoch 35/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2647.8376 - mean_absolute_error: 37.3893 - val_loss: 3124.4680 - val_mean_absolute_error: 39.5365\n",
      "Epoch 36/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2470.2976 - mean_absolute_error: 36.3289 - val_loss: 2830.4878 - val_mean_absolute_error: 37.8281\n",
      "Epoch 37/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2277.4246 - mean_absolute_error: 34.7582 - val_loss: 2666.7266 - val_mean_absolute_error: 36.7828\n",
      "Epoch 38/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2096.7507 - mean_absolute_error: 33.4248 - val_loss: 2470.7996 - val_mean_absolute_error: 35.4920\n",
      "Epoch 39/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1923.1028 - mean_absolute_error: 32.1416 - val_loss: 2279.9460 - val_mean_absolute_error: 33.9433\n",
      "Epoch 40/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1767.0505 - mean_absolute_error: 30.6905 - val_loss: 2115.1169 - val_mean_absolute_error: 32.8916\n",
      "Epoch 41/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1622.6940 - mean_absolute_error: 29.4614 - val_loss: 1964.6478 - val_mean_absolute_error: 31.9438\n",
      "Epoch 42/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1488.8934 - mean_absolute_error: 28.2339 - val_loss: 1821.0792 - val_mean_absolute_error: 30.6944\n",
      "Epoch 43/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1365.9910 - mean_absolute_error: 27.1420 - val_loss: 1694.0898 - val_mean_absolute_error: 29.3208\n",
      "Epoch 44/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1263.5852 - mean_absolute_error: 25.9893 - val_loss: 1590.6334 - val_mean_absolute_error: 28.3491\n",
      "Epoch 45/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1181.6444 - mean_absolute_error: 24.9400 - val_loss: 1480.6340 - val_mean_absolute_error: 27.0687\n",
      "Epoch 46/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 1099.2527 - mean_absolute_error: 23.7240 - val_loss: 1413.1490 - val_mean_absolute_error: 26.8095\n",
      "Epoch 47/100\n",
      "56/56 [==============================] - 0s 2ms/step - loss: 1024.0365 - mean_absolute_error: 23.2741 - val_loss: 1312.8966 - val_mean_absolute_error: 25.5783\n",
      "Epoch 48/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 974.5589 - mean_absolute_error: 22.5926 - val_loss: 1242.3925 - val_mean_absolute_error: 24.8273\n",
      "Epoch 49/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 902.5064 - mean_absolute_error: 21.5293 - val_loss: 1179.0234 - val_mean_absolute_error: 24.0050\n",
      "Epoch 50/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 851.6075 - mean_absolute_error: 20.7697 - val_loss: 1122.5936 - val_mean_absolute_error: 23.4057\n",
      "Epoch 51/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 813.7508 - mean_absolute_error: 20.3925 - val_loss: 1083.3738 - val_mean_absolute_error: 22.7014\n",
      "Epoch 52/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 773.1543 - mean_absolute_error: 19.6407 - val_loss: 1013.9885 - val_mean_absolute_error: 22.1339\n",
      "Epoch 53/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 722.6181 - mean_absolute_error: 19.0439 - val_loss: 973.0472 - val_mean_absolute_error: 21.1809\n",
      "Epoch 54/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 694.1748 - mean_absolute_error: 18.4468 - val_loss: 945.5018 - val_mean_absolute_error: 20.9031\n",
      "Epoch 55/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 660.9735 - mean_absolute_error: 17.8265 - val_loss: 881.1630 - val_mean_absolute_error: 20.3336\n",
      "Epoch 56/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 631.9750 - mean_absolute_error: 17.3619 - val_loss: 850.8969 - val_mean_absolute_error: 19.9372\n",
      "Epoch 57/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 597.9529 - mean_absolute_error: 16.8574 - val_loss: 811.9478 - val_mean_absolute_error: 19.1559\n",
      "Epoch 58/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 570.1400 - mean_absolute_error: 16.2721 - val_loss: 770.0501 - val_mean_absolute_error: 18.6503\n",
      "Epoch 59/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 539.4565 - mean_absolute_error: 15.6861 - val_loss: 740.2069 - val_mean_absolute_error: 18.0716\n",
      "Epoch 60/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 515.4504 - mean_absolute_error: 15.3594 - val_loss: 708.3535 - val_mean_absolute_error: 17.3614\n",
      "Epoch 61/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 499.1982 - mean_absolute_error: 15.0947 - val_loss: 685.6063 - val_mean_absolute_error: 16.8953\n",
      "Epoch 62/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 477.2658 - mean_absolute_error: 14.5500 - val_loss: 641.0952 - val_mean_absolute_error: 16.4600\n",
      "Epoch 63/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 455.6060 - mean_absolute_error: 14.2089 - val_loss: 619.3032 - val_mean_absolute_error: 16.0382\n",
      "Epoch 64/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 437.4860 - mean_absolute_error: 13.5667 - val_loss: 609.0177 - val_mean_absolute_error: 16.1773\n",
      "Epoch 65/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 425.9540 - mean_absolute_error: 13.3583 - val_loss: 565.8303 - val_mean_absolute_error: 15.3087\n",
      "Epoch 66/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 399.9963 - mean_absolute_error: 12.8592 - val_loss: 545.9020 - val_mean_absolute_error: 14.7424\n",
      "Epoch 67/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 385.5060 - mean_absolute_error: 12.4519 - val_loss: 524.8514 - val_mean_absolute_error: 14.2424\n",
      "Epoch 68/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 368.4361 - mean_absolute_error: 12.2206 - val_loss: 503.4949 - val_mean_absolute_error: 13.9080\n",
      "Epoch 69/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 355.6591 - mean_absolute_error: 11.9386 - val_loss: 492.7903 - val_mean_absolute_error: 13.7545\n",
      "Epoch 70/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 338.3815 - mean_absolute_error: 11.5848 - val_loss: 469.3416 - val_mean_absolute_error: 13.4877\n",
      "Epoch 71/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 340.1216 - mean_absolute_error: 11.6385 - val_loss: 458.8303 - val_mean_absolute_error: 13.7829\n",
      "Epoch 72/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 315.9507 - mean_absolute_error: 11.3292 - val_loss: 436.6569 - val_mean_absolute_error: 12.8212\n",
      "Epoch 73/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 304.6320 - mean_absolute_error: 10.9284 - val_loss: 419.0643 - val_mean_absolute_error: 12.9256\n",
      "Epoch 74/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 296.5352 - mean_absolute_error: 10.8966 - val_loss: 400.0808 - val_mean_absolute_error: 12.1685\n",
      "Epoch 75/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 287.2911 - mean_absolute_error: 10.5265 - val_loss: 393.1248 - val_mean_absolute_error: 11.8133\n",
      "Epoch 76/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 277.6201 - mean_absolute_error: 10.3789 - val_loss: 378.4101 - val_mean_absolute_error: 11.8916\n",
      "Epoch 77/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 265.9451 - mean_absolute_error: 10.1592 - val_loss: 363.0512 - val_mean_absolute_error: 11.4307\n",
      "Epoch 78/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 255.5040 - mean_absolute_error: 9.8669 - val_loss: 352.2726 - val_mean_absolute_error: 11.4597\n",
      "Epoch 79/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 247.9690 - mean_absolute_error: 9.7656 - val_loss: 339.6790 - val_mean_absolute_error: 10.9478\n",
      "Epoch 80/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 240.1246 - mean_absolute_error: 9.5203 - val_loss: 327.0750 - val_mean_absolute_error: 10.9837\n",
      "Epoch 81/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 228.5457 - mean_absolute_error: 9.3802 - val_loss: 316.7265 - val_mean_absolute_error: 10.3485\n",
      "Epoch 82/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 218.8954 - mean_absolute_error: 9.0932 - val_loss: 308.5273 - val_mean_absolute_error: 10.7565\n",
      "Epoch 83/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 216.9066 - mean_absolute_error: 9.1114 - val_loss: 296.3347 - val_mean_absolute_error: 10.1200\n",
      "Epoch 84/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 213.8757 - mean_absolute_error: 9.0485 - val_loss: 290.3912 - val_mean_absolute_error: 10.5629\n",
      "Epoch 85/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 199.8182 - mean_absolute_error: 8.8026 - val_loss: 273.4545 - val_mean_absolute_error: 9.9091\n",
      "Epoch 86/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 195.9066 - mean_absolute_error: 8.6128 - val_loss: 269.5889 - val_mean_absolute_error: 9.6979\n",
      "Epoch 87/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 189.1062 - mean_absolute_error: 8.5604 - val_loss: 264.3967 - val_mean_absolute_error: 9.6075\n",
      "Epoch 88/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 184.9591 - mean_absolute_error: 8.4408 - val_loss: 251.1330 - val_mean_absolute_error: 9.5173\n",
      "Epoch 89/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 180.9696 - mean_absolute_error: 8.2753 - val_loss: 254.3359 - val_mean_absolute_error: 9.3952\n",
      "Epoch 90/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 174.6454 - mean_absolute_error: 8.1653 - val_loss: 232.2124 - val_mean_absolute_error: 9.3542\n",
      "Epoch 91/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 165.8267 - mean_absolute_error: 8.0627 - val_loss: 239.8732 - val_mean_absolute_error: 9.1651\n",
      "Epoch 92/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 161.5656 - mean_absolute_error: 7.9436 - val_loss: 256.9140 - val_mean_absolute_error: 9.4880\n",
      "Epoch 93/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 167.8452 - mean_absolute_error: 8.1976 - val_loss: 223.9798 - val_mean_absolute_error: 8.7034\n",
      "Epoch 94/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 151.6651 - mean_absolute_error: 7.6760 - val_loss: 213.9233 - val_mean_absolute_error: 8.4075\n",
      "Epoch 95/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 147.4271 - mean_absolute_error: 7.5546 - val_loss: 206.2616 - val_mean_absolute_error: 8.3457\n",
      "Epoch 96/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 139.7100 - mean_absolute_error: 7.2803 - val_loss: 204.5311 - val_mean_absolute_error: 8.5196\n",
      "Epoch 97/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 136.6948 - mean_absolute_error: 7.2767 - val_loss: 188.1354 - val_mean_absolute_error: 8.1635\n",
      "Epoch 98/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 133.1936 - mean_absolute_error: 7.1529 - val_loss: 183.6082 - val_mean_absolute_error: 8.1459\n",
      "Epoch 99/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 135.6983 - mean_absolute_error: 7.3006 - val_loss: 185.3685 - val_mean_absolute_error: 8.0418\n",
      "Epoch 100/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 126.1125 - mean_absolute_error: 7.1078 - val_loss: 172.4787 - val_mean_absolute_error: 7.6981\n",
      "14/14 [==============================] - 0s 2ms/step - loss: 172.4787 - mean_absolute_error: 7.6981\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "56/56 [==============================] - 1s 6ms/step - loss: 183274.0000 - mean_absolute_error: 252.7893 - val_loss: 175346.7344 - val_mean_absolute_error: 251.8253\n",
      "Epoch 2/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 180376.3438 - mean_absolute_error: 250.0765 - val_loss: 169688.3594 - val_mean_absolute_error: 246.3808\n",
      "Epoch 3/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 168722.8438 - mean_absolute_error: 240.2230 - val_loss: 151026.2500 - val_mean_absolute_error: 230.0414\n",
      "Epoch 4/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 139605.2188 - mean_absolute_error: 216.1853 - val_loss: 113203.9609 - val_mean_absolute_error: 196.3312\n",
      "Epoch 5/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 93502.5078 - mean_absolute_error: 174.5381 - val_loss: 66848.7109 - val_mean_absolute_error: 145.9571\n",
      "Epoch 6/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 50574.1641 - mean_absolute_error: 123.6791 - val_loss: 33956.4453 - val_mean_absolute_error: 100.7241\n",
      "Epoch 7/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 27810.1191 - mean_absolute_error: 95.9063 - val_loss: 22337.8789 - val_mean_absolute_error: 89.3810\n",
      "Epoch 8/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 20124.4277 - mean_absolute_error: 88.5747 - val_loss: 18287.1211 - val_mean_absolute_error: 86.3104\n",
      "Epoch 9/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 16931.4785 - mean_absolute_error: 85.4403 - val_loss: 15547.7910 - val_mean_absolute_error: 81.4996\n",
      "Epoch 10/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 14679.9834 - mean_absolute_error: 80.0377 - val_loss: 13514.3848 - val_mean_absolute_error: 77.5762\n",
      "Epoch 11/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 12773.2783 - mean_absolute_error: 76.7419 - val_loss: 11765.6152 - val_mean_absolute_error: 73.0163\n",
      "Epoch 12/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 11327.8154 - mean_absolute_error: 72.7719 - val_loss: 10533.4600 - val_mean_absolute_error: 70.5984\n",
      "Epoch 13/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 10172.3506 - mean_absolute_error: 69.8233 - val_loss: 9499.6533 - val_mean_absolute_error: 68.0902\n",
      "Epoch 14/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 9185.6152 - mean_absolute_error: 67.0913 - val_loss: 8511.9697 - val_mean_absolute_error: 64.6888\n",
      "Epoch 15/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 8413.5508 - mean_absolute_error: 64.7247 - val_loss: 7824.1123 - val_mean_absolute_error: 62.3717\n",
      "Epoch 16/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 7759.7075 - mean_absolute_error: 62.6748 - val_loss: 7260.5264 - val_mean_absolute_error: 60.5247\n",
      "Epoch 17/100\n",
      "56/56 [==============================] - 0s 2ms/step - loss: 7226.9009 - mean_absolute_error: 60.5940 - val_loss: 6921.2720 - val_mean_absolute_error: 58.9560\n",
      "Epoch 18/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6815.7007 - mean_absolute_error: 58.6150 - val_loss: 6452.0425 - val_mean_absolute_error: 57.0200\n",
      "Epoch 19/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6417.1641 - mean_absolute_error: 56.7036 - val_loss: 6122.4092 - val_mean_absolute_error: 55.2837\n",
      "Epoch 20/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6071.6460 - mean_absolute_error: 55.1637 - val_loss: 5825.7134 - val_mean_absolute_error: 53.7644\n",
      "Epoch 21/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 5767.9453 - mean_absolute_error: 53.3575 - val_loss: 5553.8677 - val_mean_absolute_error: 52.1570\n",
      "Epoch 22/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 5471.9507 - mean_absolute_error: 51.7038 - val_loss: 5269.2632 - val_mean_absolute_error: 50.8315\n",
      "Epoch 23/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 5211.8711 - mean_absolute_error: 50.2950 - val_loss: 5037.7837 - val_mean_absolute_error: 49.4106\n",
      "Epoch 24/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4964.0337 - mean_absolute_error: 48.7198 - val_loss: 4831.2241 - val_mean_absolute_error: 48.0684\n",
      "Epoch 25/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4700.0625 - mean_absolute_error: 47.4170 - val_loss: 4542.1460 - val_mean_absolute_error: 46.9214\n",
      "Epoch 26/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 4461.0928 - mean_absolute_error: 45.9612 - val_loss: 4284.6509 - val_mean_absolute_error: 45.6958\n",
      "Epoch 27/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4216.8252 - mean_absolute_error: 44.8605 - val_loss: 4015.0842 - val_mean_absolute_error: 44.7305\n",
      "Epoch 28/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 3992.0261 - mean_absolute_error: 43.4569 - val_loss: 3778.5466 - val_mean_absolute_error: 43.4191\n",
      "Epoch 29/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3735.1440 - mean_absolute_error: 42.2684 - val_loss: 3558.5806 - val_mean_absolute_error: 42.1504\n",
      "Epoch 30/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3494.7996 - mean_absolute_error: 40.8282 - val_loss: 3325.7297 - val_mean_absolute_error: 40.9174\n",
      "Epoch 31/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 3250.2388 - mean_absolute_error: 39.4098 - val_loss: 3100.7341 - val_mean_absolute_error: 39.6108\n",
      "Epoch 32/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3026.7378 - mean_absolute_error: 38.4923 - val_loss: 2892.3379 - val_mean_absolute_error: 38.4370\n",
      "Epoch 33/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2789.5854 - mean_absolute_error: 36.6439 - val_loss: 2642.9194 - val_mean_absolute_error: 36.7349\n",
      "Epoch 34/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 2560.9519 - mean_absolute_error: 35.2472 - val_loss: 2434.5708 - val_mean_absolute_error: 35.6002\n",
      "Epoch 35/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2357.0364 - mean_absolute_error: 34.0372 - val_loss: 2216.6714 - val_mean_absolute_error: 34.3434\n",
      "Epoch 36/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 2164.1948 - mean_absolute_error: 32.8589 - val_loss: 2024.0807 - val_mean_absolute_error: 33.0479\n",
      "Epoch 37/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 1977.1063 - mean_absolute_error: 31.5096 - val_loss: 1846.6254 - val_mean_absolute_error: 31.8563\n",
      "Epoch 38/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 1791.8834 - mean_absolute_error: 30.3977 - val_loss: 1693.5138 - val_mean_absolute_error: 30.4228\n",
      "Epoch 39/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 1645.7078 - mean_absolute_error: 29.1858 - val_loss: 1548.3441 - val_mean_absolute_error: 29.1459\n",
      "Epoch 40/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 1491.1649 - mean_absolute_error: 27.8243 - val_loss: 1393.0609 - val_mean_absolute_error: 28.1264\n",
      "Epoch 41/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 1361.2861 - mean_absolute_error: 26.9715 - val_loss: 1279.5197 - val_mean_absolute_error: 26.8139\n",
      "Epoch 42/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 1250.2996 - mean_absolute_error: 25.7859 - val_loss: 1166.9413 - val_mean_absolute_error: 25.6832\n",
      "Epoch 43/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 1142.3875 - mean_absolute_error: 24.7765 - val_loss: 1082.3478 - val_mean_absolute_error: 24.5280\n",
      "Epoch 44/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 1056.1475 - mean_absolute_error: 23.5754 - val_loss: 986.0016 - val_mean_absolute_error: 23.7941\n",
      "Epoch 45/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 964.3429 - mean_absolute_error: 22.4589 - val_loss: 914.3765 - val_mean_absolute_error: 22.7201\n",
      "Epoch 46/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 901.6523 - mean_absolute_error: 21.7127 - val_loss: 839.0602 - val_mean_absolute_error: 21.5974\n",
      "Epoch 47/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 826.3381 - mean_absolute_error: 20.6230 - val_loss: 784.2451 - val_mean_absolute_error: 20.7443\n",
      "Epoch 48/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 771.0888 - mean_absolute_error: 19.8093 - val_loss: 720.3159 - val_mean_absolute_error: 19.9404\n",
      "Epoch 49/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 720.0300 - mean_absolute_error: 19.0463 - val_loss: 668.6501 - val_mean_absolute_error: 18.9537\n",
      "Epoch 50/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 677.0352 - mean_absolute_error: 18.1946 - val_loss: 623.4017 - val_mean_absolute_error: 18.2516\n",
      "Epoch 51/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 631.9198 - mean_absolute_error: 17.3872 - val_loss: 585.2285 - val_mean_absolute_error: 17.3997\n",
      "Epoch 52/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 595.9839 - mean_absolute_error: 16.5968 - val_loss: 548.2888 - val_mean_absolute_error: 16.6290\n",
      "Epoch 53/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 560.1558 - mean_absolute_error: 16.0102 - val_loss: 512.4351 - val_mean_absolute_error: 16.0380\n",
      "Epoch 54/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 531.2535 - mean_absolute_error: 15.3600 - val_loss: 485.3993 - val_mean_absolute_error: 15.4432\n",
      "Epoch 55/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 507.2390 - mean_absolute_error: 14.7547 - val_loss: 474.0042 - val_mean_absolute_error: 15.0639\n",
      "Epoch 56/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 486.9239 - mean_absolute_error: 14.3201 - val_loss: 436.1027 - val_mean_absolute_error: 14.4162\n",
      "Epoch 57/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 467.3752 - mean_absolute_error: 13.9365 - val_loss: 420.0320 - val_mean_absolute_error: 13.9448\n",
      "Epoch 58/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 442.5093 - mean_absolute_error: 13.3286 - val_loss: 397.8087 - val_mean_absolute_error: 13.4197\n",
      "Epoch 59/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 424.4944 - mean_absolute_error: 12.9676 - val_loss: 386.9040 - val_mean_absolute_error: 13.2622\n",
      "Epoch 60/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 411.8068 - mean_absolute_error: 12.5410 - val_loss: 376.3222 - val_mean_absolute_error: 12.9256\n",
      "Epoch 61/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 395.7431 - mean_absolute_error: 12.2477 - val_loss: 369.5167 - val_mean_absolute_error: 13.0525\n",
      "Epoch 62/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 381.6169 - mean_absolute_error: 11.8723 - val_loss: 344.2550 - val_mean_absolute_error: 12.1646\n",
      "Epoch 63/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 372.1884 - mean_absolute_error: 11.6977 - val_loss: 333.4418 - val_mean_absolute_error: 11.8271\n",
      "Epoch 64/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 366.9881 - mean_absolute_error: 11.4907 - val_loss: 318.0958 - val_mean_absolute_error: 11.4882\n",
      "Epoch 65/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 351.5345 - mean_absolute_error: 11.2309 - val_loss: 309.0754 - val_mean_absolute_error: 11.3841\n",
      "Epoch 66/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 344.0468 - mean_absolute_error: 11.0096 - val_loss: 296.2181 - val_mean_absolute_error: 11.1099\n",
      "Epoch 67/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 331.1043 - mean_absolute_error: 10.7432 - val_loss: 291.4442 - val_mean_absolute_error: 11.0226\n",
      "Epoch 68/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 324.6747 - mean_absolute_error: 10.6082 - val_loss: 275.9256 - val_mean_absolute_error: 10.6945\n",
      "Epoch 69/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 329.8396 - mean_absolute_error: 10.8724 - val_loss: 266.9981 - val_mean_absolute_error: 10.4593\n",
      "Epoch 70/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 306.5399 - mean_absolute_error: 10.3464 - val_loss: 256.4845 - val_mean_absolute_error: 10.5027\n",
      "Epoch 71/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 295.0837 - mean_absolute_error: 10.2562 - val_loss: 258.4981 - val_mean_absolute_error: 10.1330\n",
      "Epoch 72/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 289.2895 - mean_absolute_error: 9.9356 - val_loss: 237.7460 - val_mean_absolute_error: 9.9778\n",
      "Epoch 73/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 281.9526 - mean_absolute_error: 9.9284 - val_loss: 233.8318 - val_mean_absolute_error: 9.7997\n",
      "Epoch 74/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 277.8488 - mean_absolute_error: 9.9110 - val_loss: 224.9845 - val_mean_absolute_error: 9.7074\n",
      "Epoch 75/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 270.5373 - mean_absolute_error: 9.7010 - val_loss: 223.9702 - val_mean_absolute_error: 9.7676\n",
      "Epoch 76/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 269.2875 - mean_absolute_error: 9.8025 - val_loss: 218.1053 - val_mean_absolute_error: 9.6711\n",
      "Epoch 77/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 251.7346 - mean_absolute_error: 9.4007 - val_loss: 205.6312 - val_mean_absolute_error: 9.3664\n",
      "Epoch 78/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 247.3902 - mean_absolute_error: 9.3269 - val_loss: 196.9424 - val_mean_absolute_error: 9.1248\n",
      "Epoch 79/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 239.5098 - mean_absolute_error: 9.2178 - val_loss: 198.0896 - val_mean_absolute_error: 9.0736\n",
      "Epoch 80/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 232.9931 - mean_absolute_error: 9.1849 - val_loss: 190.6341 - val_mean_absolute_error: 8.8017\n",
      "Epoch 81/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 237.2298 - mean_absolute_error: 9.1670 - val_loss: 177.8397 - val_mean_absolute_error: 8.8077\n",
      "Epoch 82/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 219.4111 - mean_absolute_error: 8.9223 - val_loss: 173.6512 - val_mean_absolute_error: 8.6427\n",
      "Epoch 83/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 221.7347 - mean_absolute_error: 9.0279 - val_loss: 167.0654 - val_mean_absolute_error: 8.5298\n",
      "Epoch 84/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 207.6500 - mean_absolute_error: 8.7014 - val_loss: 166.3939 - val_mean_absolute_error: 8.7329\n",
      "Epoch 85/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 205.9849 - mean_absolute_error: 8.7690 - val_loss: 171.2794 - val_mean_absolute_error: 8.7587\n",
      "Epoch 86/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 204.9212 - mean_absolute_error: 8.6783 - val_loss: 155.3930 - val_mean_absolute_error: 8.4495\n",
      "Epoch 87/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 194.4148 - mean_absolute_error: 8.5637 - val_loss: 148.8057 - val_mean_absolute_error: 8.4392\n",
      "Epoch 88/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 190.2188 - mean_absolute_error: 8.3898 - val_loss: 148.9991 - val_mean_absolute_error: 8.4688\n",
      "Epoch 89/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 184.0035 - mean_absolute_error: 8.3467 - val_loss: 140.7850 - val_mean_absolute_error: 8.0534\n",
      "Epoch 90/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 177.3874 - mean_absolute_error: 8.2342 - val_loss: 136.8503 - val_mean_absolute_error: 7.7837\n",
      "Epoch 91/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 179.6339 - mean_absolute_error: 8.1854 - val_loss: 146.7586 - val_mean_absolute_error: 8.5235\n",
      "Epoch 92/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 181.7955 - mean_absolute_error: 8.5332 - val_loss: 131.3627 - val_mean_absolute_error: 8.0598\n",
      "Epoch 93/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 167.6617 - mean_absolute_error: 8.0193 - val_loss: 123.6272 - val_mean_absolute_error: 7.5924\n",
      "Epoch 94/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 162.0275 - mean_absolute_error: 7.8205 - val_loss: 119.1860 - val_mean_absolute_error: 7.3371\n",
      "Epoch 95/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 160.1790 - mean_absolute_error: 7.6906 - val_loss: 130.4214 - val_mean_absolute_error: 8.0431\n",
      "Epoch 96/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 154.1808 - mean_absolute_error: 7.7884 - val_loss: 112.7949 - val_mean_absolute_error: 7.3782\n",
      "Epoch 97/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 148.2537 - mean_absolute_error: 7.6080 - val_loss: 111.2022 - val_mean_absolute_error: 7.2299\n",
      "Epoch 98/100\n",
      "56/56 [==============================] - 0s 2ms/step - loss: 145.8075 - mean_absolute_error: 7.4772 - val_loss: 106.4835 - val_mean_absolute_error: 7.3422\n",
      "Epoch 99/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 135.2347 - mean_absolute_error: 7.2345 - val_loss: 102.5084 - val_mean_absolute_error: 7.0932\n",
      "Epoch 100/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 137.1766 - mean_absolute_error: 7.3297 - val_loss: 100.2040 - val_mean_absolute_error: 7.0358\n",
      "14/14 [==============================] - 0s 3ms/step - loss: 100.2040 - mean_absolute_error: 7.0358\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "56/56 [==============================] - 1s 6ms/step - loss: 180750.6719 - mean_absolute_error: 254.7992 - val_loss: 182907.4062 - val_mean_absolute_error: 241.8209\n",
      "Epoch 2/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 176938.5938 - mean_absolute_error: 251.4189 - val_loss: 175893.4844 - val_mean_absolute_error: 236.0505\n",
      "Epoch 3/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 163890.2656 - mean_absolute_error: 240.4818 - val_loss: 155554.7344 - val_mean_absolute_error: 219.9093\n",
      "Epoch 4/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 134112.1562 - mean_absolute_error: 215.6025 - val_loss: 117010.2734 - val_mean_absolute_error: 188.2174\n",
      "Epoch 5/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 89992.3672 - mean_absolute_error: 172.3879 - val_loss: 69851.3359 - val_mean_absolute_error: 141.5980\n",
      "Epoch 6/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 48563.1797 - mean_absolute_error: 122.2773 - val_loss: 35635.6523 - val_mean_absolute_error: 100.5720\n",
      "Epoch 7/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 26658.0137 - mean_absolute_error: 93.8717 - val_loss: 23162.8496 - val_mean_absolute_error: 89.5772\n",
      "Epoch 8/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 20220.1621 - mean_absolute_error: 89.2907 - val_loss: 19292.5625 - val_mean_absolute_error: 86.7809\n",
      "Epoch 9/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 17712.8398 - mean_absolute_error: 87.1617 - val_loss: 17070.7520 - val_mean_absolute_error: 83.2547\n",
      "Epoch 10/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 15880.3545 - mean_absolute_error: 83.5756 - val_loss: 15424.2998 - val_mean_absolute_error: 79.7001\n",
      "Epoch 11/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 14427.6045 - mean_absolute_error: 80.1483 - val_loss: 13973.4307 - val_mean_absolute_error: 76.6639\n",
      "Epoch 12/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 13163.5840 - mean_absolute_error: 76.9035 - val_loss: 12694.3350 - val_mean_absolute_error: 73.8541\n",
      "Epoch 13/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 12079.1494 - mean_absolute_error: 74.4369 - val_loss: 11662.3613 - val_mean_absolute_error: 71.2927\n",
      "Epoch 14/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 11182.3447 - mean_absolute_error: 72.0893 - val_loss: 10743.0527 - val_mean_absolute_error: 68.8222\n",
      "Epoch 15/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 10379.6172 - mean_absolute_error: 70.0632 - val_loss: 10039.3506 - val_mean_absolute_error: 66.9431\n",
      "Epoch 16/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 9714.5566 - mean_absolute_error: 68.0751 - val_loss: 9240.3105 - val_mean_absolute_error: 64.9669\n",
      "Epoch 17/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 9122.3379 - mean_absolute_error: 66.5006 - val_loss: 8689.0449 - val_mean_absolute_error: 63.4105\n",
      "Epoch 18/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 8629.4209 - mean_absolute_error: 65.1375 - val_loss: 8222.8184 - val_mean_absolute_error: 62.2698\n",
      "Epoch 19/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 8148.0825 - mean_absolute_error: 63.5284 - val_loss: 7747.0234 - val_mean_absolute_error: 60.5277\n",
      "Epoch 20/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 7767.9209 - mean_absolute_error: 62.2536 - val_loss: 7456.4678 - val_mean_absolute_error: 59.6715\n",
      "Epoch 21/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 7427.4639 - mean_absolute_error: 61.0276 - val_loss: 7039.3755 - val_mean_absolute_error: 57.8326\n",
      "Epoch 22/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 7128.2915 - mean_absolute_error: 59.7304 - val_loss: 6805.2544 - val_mean_absolute_error: 57.0082\n",
      "Epoch 23/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6880.9766 - mean_absolute_error: 58.7361 - val_loss: 6498.6123 - val_mean_absolute_error: 55.4428\n",
      "Epoch 24/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6626.4502 - mean_absolute_error: 57.4568 - val_loss: 6286.8022 - val_mean_absolute_error: 54.5367\n",
      "Epoch 25/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6392.3130 - mean_absolute_error: 56.4163 - val_loss: 6006.7837 - val_mean_absolute_error: 53.1040\n",
      "Epoch 26/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6198.3931 - mean_absolute_error: 55.4114 - val_loss: 5812.2290 - val_mean_absolute_error: 52.2308\n",
      "Epoch 27/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 6009.1689 - mean_absolute_error: 54.6037 - val_loss: 5614.5981 - val_mean_absolute_error: 51.2474\n",
      "Epoch 28/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 5791.8071 - mean_absolute_error: 53.3329 - val_loss: 5457.0659 - val_mean_absolute_error: 50.5762\n",
      "Epoch 29/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 5611.1885 - mean_absolute_error: 52.6285 - val_loss: 5226.7588 - val_mean_absolute_error: 49.4454\n",
      "Epoch 30/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 5383.1128 - mean_absolute_error: 51.3989 - val_loss: 5128.6099 - val_mean_absolute_error: 49.1295\n",
      "Epoch 31/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 5197.3618 - mean_absolute_error: 50.5644 - val_loss: 4840.1221 - val_mean_absolute_error: 47.4433\n",
      "Epoch 32/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4998.7275 - mean_absolute_error: 49.5861 - val_loss: 4656.0791 - val_mean_absolute_error: 46.6572\n",
      "Epoch 33/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4809.8335 - mean_absolute_error: 48.5400 - val_loss: 4463.7124 - val_mean_absolute_error: 45.8827\n",
      "Epoch 34/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 4607.4961 - mean_absolute_error: 47.6646 - val_loss: 4277.2437 - val_mean_absolute_error: 45.1629\n",
      "Epoch 35/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4428.8115 - mean_absolute_error: 46.6555 - val_loss: 4126.3809 - val_mean_absolute_error: 44.5691\n",
      "Epoch 36/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4215.1885 - mean_absolute_error: 45.4366 - val_loss: 3909.8518 - val_mean_absolute_error: 43.4564\n",
      "Epoch 37/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 4031.0234 - mean_absolute_error: 45.0306 - val_loss: 3714.3794 - val_mean_absolute_error: 42.4057\n",
      "Epoch 38/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3853.3821 - mean_absolute_error: 43.9947 - val_loss: 3522.8318 - val_mean_absolute_error: 41.3302\n",
      "Epoch 39/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3635.5320 - mean_absolute_error: 42.8176 - val_loss: 3335.5730 - val_mean_absolute_error: 40.5197\n",
      "Epoch 40/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 3452.2771 - mean_absolute_error: 41.7978 - val_loss: 3173.8652 - val_mean_absolute_error: 39.5294\n",
      "Epoch 41/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 3258.3521 - mean_absolute_error: 40.6871 - val_loss: 3005.0649 - val_mean_absolute_error: 38.7565\n",
      "Epoch 42/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 3092.6272 - mean_absolute_error: 39.8929 - val_loss: 2836.8511 - val_mean_absolute_error: 37.8458\n",
      "Epoch 43/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2945.6028 - mean_absolute_error: 38.9241 - val_loss: 2674.1543 - val_mean_absolute_error: 36.6517\n",
      "Epoch 44/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2744.2947 - mean_absolute_error: 37.5054 - val_loss: 2537.0239 - val_mean_absolute_error: 35.9861\n",
      "Epoch 45/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2587.7500 - mean_absolute_error: 36.5203 - val_loss: 2350.2258 - val_mean_absolute_error: 34.6086\n",
      "Epoch 46/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2425.5728 - mean_absolute_error: 35.2577 - val_loss: 2192.6331 - val_mean_absolute_error: 33.4483\n",
      "Epoch 47/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 2258.1653 - mean_absolute_error: 33.8888 - val_loss: 2043.4297 - val_mean_absolute_error: 32.2741\n",
      "Epoch 48/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 2104.1577 - mean_absolute_error: 32.7561 - val_loss: 1910.1263 - val_mean_absolute_error: 31.2720\n",
      "Epoch 49/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1971.7615 - mean_absolute_error: 31.6614 - val_loss: 1782.2590 - val_mean_absolute_error: 30.2030\n",
      "Epoch 50/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1824.8055 - mean_absolute_error: 30.7089 - val_loss: 1655.4426 - val_mean_absolute_error: 29.0292\n",
      "Epoch 51/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1702.9893 - mean_absolute_error: 29.3283 - val_loss: 1555.1799 - val_mean_absolute_error: 28.1637\n",
      "Epoch 52/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 1584.7589 - mean_absolute_error: 28.3380 - val_loss: 1430.3977 - val_mean_absolute_error: 26.9210\n",
      "Epoch 53/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 1478.0714 - mean_absolute_error: 27.2648 - val_loss: 1331.0531 - val_mean_absolute_error: 25.9089\n",
      "Epoch 54/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 1366.1693 - mean_absolute_error: 26.1068 - val_loss: 1240.6920 - val_mean_absolute_error: 24.8828\n",
      "Epoch 55/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 1277.9281 - mean_absolute_error: 25.1947 - val_loss: 1167.6368 - val_mean_absolute_error: 24.1085\n",
      "Epoch 56/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 1197.4675 - mean_absolute_error: 24.3211 - val_loss: 1079.3168 - val_mean_absolute_error: 23.2200\n",
      "Epoch 57/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 1101.7294 - mean_absolute_error: 23.3836 - val_loss: 1006.9901 - val_mean_absolute_error: 22.2887\n",
      "Epoch 58/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 1054.4344 - mean_absolute_error: 22.5685 - val_loss: 941.3998 - val_mean_absolute_error: 21.5025\n",
      "Epoch 59/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 961.6173 - mean_absolute_error: 21.4957 - val_loss: 888.3426 - val_mean_absolute_error: 20.7534\n",
      "Epoch 60/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 896.3527 - mean_absolute_error: 20.6996 - val_loss: 831.1582 - val_mean_absolute_error: 20.0318\n",
      "Epoch 61/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 842.3664 - mean_absolute_error: 19.8741 - val_loss: 777.4506 - val_mean_absolute_error: 19.1350\n",
      "Epoch 62/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 797.3773 - mean_absolute_error: 19.3764 - val_loss: 744.7031 - val_mean_absolute_error: 18.8613\n",
      "Epoch 63/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 741.5898 - mean_absolute_error: 18.4559 - val_loss: 704.8445 - val_mean_absolute_error: 18.0931\n",
      "Epoch 64/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 715.9036 - mean_absolute_error: 18.0403 - val_loss: 678.1102 - val_mean_absolute_error: 17.8969\n",
      "Epoch 65/100\n",
      "56/56 [==============================] - 0s 5ms/step - loss: 673.5533 - mean_absolute_error: 17.5477 - val_loss: 617.9473 - val_mean_absolute_error: 16.7691\n",
      "Epoch 66/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 633.0731 - mean_absolute_error: 16.7814 - val_loss: 590.9635 - val_mean_absolute_error: 16.2834\n",
      "Epoch 67/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 604.3007 - mean_absolute_error: 16.3055 - val_loss: 568.1221 - val_mean_absolute_error: 16.0326\n",
      "Epoch 68/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 578.7650 - mean_absolute_error: 15.9884 - val_loss: 544.7078 - val_mean_absolute_error: 15.5910\n",
      "Epoch 69/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 551.2849 - mean_absolute_error: 15.4371 - val_loss: 517.9295 - val_mean_absolute_error: 14.8900\n",
      "Epoch 70/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 522.3891 - mean_absolute_error: 14.9616 - val_loss: 493.6258 - val_mean_absolute_error: 14.4355\n",
      "Epoch 71/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 502.5930 - mean_absolute_error: 14.6573 - val_loss: 472.7478 - val_mean_absolute_error: 14.3048\n",
      "Epoch 72/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 483.3557 - mean_absolute_error: 14.2860 - val_loss: 458.5074 - val_mean_absolute_error: 13.7781\n",
      "Epoch 73/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 465.1613 - mean_absolute_error: 13.8759 - val_loss: 443.3242 - val_mean_absolute_error: 13.5493\n",
      "Epoch 74/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 454.7159 - mean_absolute_error: 13.7747 - val_loss: 435.7385 - val_mean_absolute_error: 13.4633\n",
      "Epoch 75/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 439.4014 - mean_absolute_error: 13.5478 - val_loss: 409.9875 - val_mean_absolute_error: 12.8795\n",
      "Epoch 76/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 435.9754 - mean_absolute_error: 13.4358 - val_loss: 396.9929 - val_mean_absolute_error: 12.6124\n",
      "Epoch 77/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 416.9220 - mean_absolute_error: 13.1265 - val_loss: 394.1194 - val_mean_absolute_error: 12.6685\n",
      "Epoch 78/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 392.1724 - mean_absolute_error: 12.5131 - val_loss: 392.5567 - val_mean_absolute_error: 12.3867\n",
      "Epoch 79/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 378.9807 - mean_absolute_error: 12.1993 - val_loss: 360.1549 - val_mean_absolute_error: 11.6999\n",
      "Epoch 80/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 365.0581 - mean_absolute_error: 11.9988 - val_loss: 346.8607 - val_mean_absolute_error: 11.6640\n",
      "Epoch 81/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 353.8907 - mean_absolute_error: 11.8422 - val_loss: 335.0567 - val_mean_absolute_error: 11.3805\n",
      "Epoch 82/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 345.1538 - mean_absolute_error: 11.5761 - val_loss: 330.5276 - val_mean_absolute_error: 11.1915\n",
      "Epoch 83/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 339.7267 - mean_absolute_error: 11.6361 - val_loss: 315.9704 - val_mean_absolute_error: 11.0884\n",
      "Epoch 84/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 320.1552 - mean_absolute_error: 11.2703 - val_loss: 304.1887 - val_mean_absolute_error: 10.8650\n",
      "Epoch 85/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 308.9110 - mean_absolute_error: 11.0895 - val_loss: 297.8201 - val_mean_absolute_error: 10.7598\n",
      "Epoch 86/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 300.1417 - mean_absolute_error: 10.9076 - val_loss: 298.2471 - val_mean_absolute_error: 10.6865\n",
      "Epoch 87/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 289.0030 - mean_absolute_error: 10.5132 - val_loss: 279.8453 - val_mean_absolute_error: 10.4646\n",
      "Epoch 88/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 285.1175 - mean_absolute_error: 10.6019 - val_loss: 285.9020 - val_mean_absolute_error: 10.5471\n",
      "Epoch 89/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 272.8379 - mean_absolute_error: 10.2789 - val_loss: 266.9662 - val_mean_absolute_error: 10.1028\n",
      "Epoch 90/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 266.7003 - mean_absolute_error: 10.2283 - val_loss: 260.5398 - val_mean_absolute_error: 10.1564\n",
      "Epoch 91/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 259.2868 - mean_absolute_error: 10.1626 - val_loss: 251.4475 - val_mean_absolute_error: 9.9884\n",
      "Epoch 92/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 256.3925 - mean_absolute_error: 10.0323 - val_loss: 246.1094 - val_mean_absolute_error: 9.7327\n",
      "Epoch 93/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 248.9517 - mean_absolute_error: 9.8653 - val_loss: 238.6587 - val_mean_absolute_error: 9.7937\n",
      "Epoch 94/100\n",
      "56/56 [==============================] - 0s 4ms/step - loss: 242.0823 - mean_absolute_error: 9.6968 - val_loss: 231.3871 - val_mean_absolute_error: 9.5786\n",
      "Epoch 95/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 232.3555 - mean_absolute_error: 9.5297 - val_loss: 225.3966 - val_mean_absolute_error: 9.4103\n",
      "Epoch 96/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 227.0246 - mean_absolute_error: 9.5129 - val_loss: 222.3061 - val_mean_absolute_error: 9.3127\n",
      "Epoch 97/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 219.7364 - mean_absolute_error: 9.4394 - val_loss: 217.7670 - val_mean_absolute_error: 9.2724\n",
      "Epoch 98/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 207.3158 - mean_absolute_error: 9.0319 - val_loss: 199.1853 - val_mean_absolute_error: 8.8284\n",
      "Epoch 99/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 192.5890 - mean_absolute_error: 8.5953 - val_loss: 182.7841 - val_mean_absolute_error: 8.3272\n",
      "Epoch 100/100\n",
      "56/56 [==============================] - 0s 3ms/step - loss: 175.2085 - mean_absolute_error: 8.0443 - val_loss: 178.0065 - val_mean_absolute_error: 8.4225\n",
      "14/14 [==============================] - 0s 2ms/step - loss: 178.0065 - mean_absolute_error: 8.4225\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean MAE across all folds: 6.947884464263916\n"
     ]
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from sklearn.model_selection import KFold\n",
    "\n",
    "k = 5\n",
    "mae_scores = []\n",
    "\n",
    "# initialize KFold cross-validator\n",
    "kf = KFold(n_splits=k, shuffle=True, random_state=42)\n",
    "\n",
    "X = np.array(X)\n",
    "y = np.array(y)\n",
    "\n",
    "# create a list to store MAE values for each fold\n",
    "mae_scores = []\n",
    "\n",
    "# initialize a counter for fold number\n",
    "fold_number = 1\n",
    "\n",
    "# initialize lists to store training and validation losses\n",
    "training_losses = []\n",
    "validation_losses = []\n",
    "\n",
    "# iterate through each fold\n",
    "for train_index, val_index in kf.split(X):\n",
    "    X_train_normalized, X_val = X[train_index], X[val_index]\n",
    "    y_train, y_val = y[train_index], y[val_index]\n",
    "\n",
    "    # create and compile your model\n",
    "    my_model = tf.keras.Sequential([\n",
    "        tf.keras.layers.Dense(64, activation='relu', input_shape=(X_train.shape[1],)),\n",
    "        tf.keras.layers.Dense(32, activation='relu'),\n",
    "        tf.keras.layers.Dense(1)\n",
    "    ])\n",
    "\n",
    "    my_model.compile(loss='mean_squared_error', optimizer='adam', metrics=['mean_absolute_error'])\n",
    "\n",
    "    # train the model on the current fold\n",
    "    history = my_model.fit(X_train_normalized, y_train, epochs=100, batch_size=32, validation_data=(X_val, y_val))\n",
    "\n",
    "    # evaluate the model on the validation data for this fold\n",
    "    _, mae = my_model.evaluate(X_val, y_val)\n",
    "    mae_scores.append(mae)\n",
    "\n",
    "    # append training and validation loss values to the respective lists\n",
    "    training_losses.append(history.history['loss'])\n",
    "    validation_losses.append(history.history['val_loss'])\n",
    "\n",
    "    # plot training and validation loss for this fold\n",
    "    plt.figure()\n",
    "    plt.plot(history.history['loss'], label='Training Loss')\n",
    "    plt.plot(history.history['val_loss'], label='Validation Loss')\n",
    "    plt.title(f'Fold {fold_number} Losses')\n",
    "    plt.xlabel('Epoch')\n",
    "    plt.ylabel('Loss')\n",
    "    plt.legend()\n",
    "    plt.show()\n",
    "\n",
    "    fold_number += 1\n",
    "\n",
    "# calculate and print the mean MAE across all folds\n",
    "mean_mae = np.mean(mae_scores)\n",
    "print(\"Mean MAE across all folds:\", mean_mae)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Vl3E3Pqh8elB"
   },
   "source": [
    "<h4> 5.1.3 Try out model</h4>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Fowlq_ii8pe9",
    "outputId": "de059c04-e469-4ee6-a818-77cc0145bc85"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14/14 [==============================] - 0s 2ms/step\n"
     ]
    }
   ],
   "source": [
    "y_pred = my_model.predict(X_test_normalized)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "08Lu7EsS8qqU",
    "outputId": "c9523c14-1172-49f2-a13c-aaf9e7b0ceb4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 257.1142127216899\n",
      "Mean Squared Error: 82560.22388034024\n",
      "Root Mean Squared Error: 287.3329495208307\n",
      "R-squared (R2) Score: 0.30298314530006176\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score\n",
    "\n",
    "mae = mean_absolute_error(y_test, y_pred)\n",
    "mse = mean_squared_error(y_test, y_pred)\n",
    "rmse = np.sqrt(mse)\n",
    "r2 = r2_score(y_test, y_pred)\n",
    "\n",
    "print(f'Mean Absolute Error: {mae}')\n",
    "print(f'Mean Squared Error: {mse}')\n",
    "print(f'Root Mean Squared Error: {rmse}')\n",
    "print(f'R-squared (R2) Score: {r2}')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 564
    },
    "id": "qhhfH-xP8wAX",
    "outputId": "ec26f8a1-a63a-479b-a734-eb4b08956849"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsYAAAIjCAYAAAAA4d2EAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB1wklEQVR4nO3de5yMdf/H8ffseRc7a7EHYi05rWMIm6icc0jRXUQJqUR3pdxy/yqHktJ95+5IB9EdOigl0naLDmIdcl5bQovKLrHtLml37cz1+2ObydjTzOzs7On1fDz28bDX9Z1rvjOGffv6XJ+vyTAMQwAAAEA151PeEwAAAAAqAoIxAAAAIIIxAAAAIIlgDAAAAEgiGAMAAACSCMYAAACAJIIxAAAAIIlgDAAAAEgiGAMAAACSCMYA4MBkMmnmzJnlPY0KaebMmTKZTA7HGjdurNtvv718JlSIwuYIAM4iGAMoMy+//LJMJpO6du3q9jWOHz+umTNnavfu3Z6bWCVlMpnsXz4+Pqpfv7769eunL7/8sryn5pKK9Hv65ZdfatiwYYqKilJAQIAiIiI0ZMgQrVy50j7myJEjMplM+te//uXw2JMnT8rPz0+jR48u8vpnzpxRcHCwhg0bVmavAYDnEIwBlJlly5apcePG2rZtmw4dOuTWNY4fP65Zs2ZViBBVEfTt21dvvfWW3nzzTd19993au3evevXqpU8//bRc5nPgwAG99tprLj2movyezpgxQ9dcc42SkpJ01113aeHChZo6darOnj2r4cOHa/ny5cU+PiIiQn379tWqVat07ty5QsesXLlS2dnZxYZnABWHX3lPAEDVlJKSos2bN2vlypW66667tGzZMs2YMaO8p1XpNW/e3CFk3XDDDWrXrp3+85//6Nprry30MdnZ2QoICJCPj+fXQgIDAz1+TW94//33NXv2bN14441avny5/P397eemTp2qzz77TOfPny/xOqNGjVJCQoI+/vhjjRgxosD55cuXy2w2a9CgQR6dP4CywYoxgDKxbNky1a5dW4MGDdKNN96oZcuWFTouIyNDDzzwgBo3bqzAwEBdcskluu2223Tq1Cl9+eWXuvzyyyVJY8eOtZcRLFmyRFLR9a1XX321rr76avv3ubm5euyxx9SpUyeZzWbVqFFDPXr00BdffOHy6zpx4oT8/Pw0a9asAucOHDggk8mkF198UZJ0/vx5zZo1S82aNVNQUJDq1KmjK6+8UuvWrXP5eYvStm1b1a1bVykpKZLySwNMJpPeeecdPfLII2rQoIFCQkKUlZUlSdq6dasGDBggs9mskJAQXXXVVdq0aVOB637zzTe6/PLLFRQUpKZNm+qVV14p9PkL+z0oze9pWcyxMI8++qjCw8P1xhtvOIRim/79+2vw4MElXueGG25QjRo1Cl1dPnnypNavX68bb7yx0v4DAqhuWDEGUCaWLVumYcOGKSAgQCNHjtSCBQu0fft2eyiSpLNnz6pHjx767rvvNG7cOHXs2FGnTp3Sxx9/rJ9//lmtWrXS7Nmz9dhjj+nOO+9Ujx49JElXXHGFS3PJysrS66+/rpEjR2rChAk6c+aMFi1apP79+2vbtm3q0KGD09eKjIzUVVddpffee6/ACvi7774rX19f/e1vf5OUfyPY3Llzdccdd6hLly7KysrSt99+q507d6pv374uvYai/Pbbb/rtt9906aWXOhx//PHHFRAQoIceekg5OTkKCAjQhg0bdO2116pTp06aMWOGfHx8tHjxYvXq1UsbN25Uly5dJEn79u1Tv379VK9ePc2cOVN5eXmaMWOGIiMjS5xPaX9PvTHHgwcP6vvvv9e4ceNUq1Ytl97vi9WoUUNDhw7V+++/r/T0dIWHh9vPvfvuu7JYLBo1alSpngOAFxkA4GHffvutIclYt26dYRiGYbVajUsuucS47777HMY99thjhiRj5cqVBa5htVoNwzCM7du3G5KMxYsXFxgTExNjjBkzpsDxq666yrjqqqvs3+fl5Rk5OTkOY3777TcjMjLSGDdunMNxScaMGTOKfX2vvPKKIcnYt2+fw/G4uDijV69e9u/bt29vDBo0qNhruUKSMX78eOPXX381Tp48aWzdutXo3bu3Icn497//bRiGYXzxxReGJKNJkybGuXPn7I+1Wq1Gs2bNjP79+9vfW8MwjHPnzhmxsbFG37597ceuv/56IygoyDh69Kj9WHJysuHr62tc/GPj4t+D0vyeltUcL7Zq1SpDkjF//vxix9mkpKQYkoxnnnmm0POffPKJIcl45ZVXHI5369bNaNCggWGxWJx6HgDlj1IKAB63bNkyRUZG6pprrpGU303h5ptv1jvvvCOLxWIf98EHH6h9+/a64YYbClzDky23fH19FRAQIEmyWq1KT09XXl6eOnfurJ07d7p8vWHDhsnPz0/vvvuu/VhSUpKSk5N1880324+FhYVp//79OnjwYOlfxJ8WLVqkevXqKSIiQl27dtWmTZs0ZcoU3X///Q7jxowZo+DgYPv3u3fv1sGDB3XLLbfo9OnTOnXqlE6dOqXff/9dvXv31tdffy2r1SqLxaLPPvtM119/vRo1amR/fKtWrdS/f/8S51ea31NvzdFWVlLa1WIb28r1heUUKSkp2rJli0aOHFkmtd0AygZ/WgF4lMVi0TvvvKNrrrlGKSkpOnTokA4dOqSuXbvqxIkTWr9+vX3s4cOH1aZNG6/M680331S7du3stb716tXTJ598oszMTJevVbduXfXu3Vvvvfee/di7774rPz8/h7Zcs2fPVkZGhpo3b662bdtq6tSp2rt3b6lex9ChQ7Vu3Tp9/vnn2rp1q06dOqV///vfBcJXbGysw/e2cD5mzBjVq1fP4ev1119XTk6OMjMz9euvv+qPP/5Qs2bNCjx3ixYtSpxfaX5PvTXH0NBQSfmt1DzBz89PN998szZu3KhffvlFkuwhmTIKoHKhxhiAR23YsEGpqal655139M477xQ4v2zZMvXr188jz1XUCqTFYpGvr6/9+6VLl+r222/X9ddfr6lTpyoiIkK+vr6aO3euDh8+7NZzjxgxQmPHjtXu3bvVoUMHvffee+rdu7fq1q1rH9OzZ08dPnxYq1at0v/+9z+9/vrrmj9/vhYuXKg77rjDree95JJL1KdPnxLHXbhaLOWvlEvSM888U2RNdc2aNZWTk+PWvDzBW3Ns2bKlpPw6ZU8ZPXq0XnzxRb399tt66KGH9PbbbysuLs6l+nUA5Y9gDMCjli1bpoiICL300ksFzq1cuVIffvihFi5cqODgYDVt2lRJSUnFXq+4/36vXbu2MjIyChw/evSomjRpYv/+/fffV5MmTbRy5UqH65Wmfdz111+vu+66y15O8cMPP2j69OkFxoWHh2vs2LEaO3aszp49q549e2rmzJluB2N3NW3aVFL+amlxwbpevXoKDg4utPzjwIEDTj2Pu7+n3ppj8+bN1aJFC61atUrPPfecatasWeJjStK1a1c1bdpUy5cvV9++fbV//37NmTOn1NcF4F2UUgDwmD/++EMrV67U4MGDdeONNxb4mjx5ss6cOaOPP/5YkjR8+HDt2bNHH374YYFrGYYhKf+uf0mFBuCmTZtqy5Ytys3NtR9bs2aNfvrpJ4dxttVj2zWl/JZgiYmJbr/WsLAw9e/fX++9957eeecdBQQE6Prrr3cYc/r0aYfva9asqUsvvdRhxTMzM1Pff/+9WyUdrujUqZOaNm2qf/3rXzp79myB87/++quk/Peqf//++uijj3Ts2DH7+e+++06fffZZic9Tmt9Tb81RkmbNmqXTp0/rjjvuUF5eXoHz//vf/7RmzRqnrmUzatQo7dq1SzNmzJDJZNItt9zi0uMBlD9WjAF4zMcff6wzZ87ouuuuK/R8t27dVK9ePS1btkw333yzpk6dqvfff19/+9vfNG7cOHXq1Enp6en6+OOPtXDhQrVv315NmzZVWFiYFi5cqFq1aqlGjRrq2rWrYmNjdccdd+j999/XgAEDdNNNN+nw4cNaunSpfeXRZvDgwVq5cqVuuOEGDRo0SCkpKVq4cKHi4uIKDWDOuvnmmzV69Gi9/PLL6t+/v8LCwhzOx8XF6eqrr1anTp0UHh6ub7/9Vu+//74mT55sH/Phhx9q7NixWrx4caE9mT3Fx8dHr7/+uq699lq1bt1aY8eOVYMGDfTLL7/oiy++UGhoqFavXi0pPzQmJCSoR48euueee5SXl6cXXnhBrVu3LrFGurS/p96Yo5T/e7dv3z7NmTNHu3bt0siRIxUTE6PTp08rISFB69evL9CbeP369crOzi5wreuvv15t2rTR6NGjNXv2bK1atUrdu3dX48aNnfzdAVBhlHNXDABVyJAhQ4ygoCDj999/L3LM7bffbvj7+xunTp0yDMMwTp8+bUyePNlo0KCBERAQYFxyySXGmDFj7OcNI7+9VlxcnOHn51egzde///1vo0GDBkZgYKDRvXt349tvvy3Qrs1qtRpPPvmkERMTYwQGBhqXXXaZsWbNGmPMmDFGTEyMw/zkRLs2m6ysLCM4ONiQZCxdurTA+SeeeMLo0qWLERYWZgQHBxstW7Y05syZY+Tm5trHLF68uMh2dBeTZEyaNKnYMbZ2bStWrCj0/K5du4xhw4YZderUMQIDA42YmBjjpptuMtavX+8w7quvvjI6depkBAQEGE2aNDEWLlxozJgxo8R2bYZR+t9TT8+xOOvXrzeGDh1qREREGH5+fka9evWMIUOGGKtWrbKPsbVrK+rrrbfeso+9/PLLDUnGyy+/7PQcAFQcJsO44P8WAQAAgGqKGmMAAABABGMAAABAEsEYAAAAkEQwBgAAACQRjAEAAABJBGMAAABAEht8SJKsVquOHz+uWrVqFbv9LAAAAMqHYRg6c+aM6tevLx+fslnbJRhLOn78uBo2bFje0wAAAEAJfvrpJ11yySVlcm2CsaRatWpJyn+jQ0NDy3k2AAAAuFhWVpYaNmxoz21lgWAs2csnQkNDCcYAAAAVWFmWvXLzHQAAACCCMQAAACCJYAwAAABIIhgDAAAAkgjGAAAAgCSCMQAAACCJYAwAAABIIhgDAAAAkgjGAAAAgCSCMQAAACCJYAwAAABIIhgDAAAAkgjGAAAAgCTJr7wnAAAAgMrDYjW0LSVdJ89kK6JWkLrEhsvXx1Te0/IIgjEAAACckpCUqlmrk5WamW0/Fm0O0owhcRrQJrocZ+YZlFIAAACgRAlJqZq4dKdDKJaktMxsTVy6UwlJqeU0M88hGAMAAKBYFquhWauTZRRyznZs1upkWayFjag8CMYAAAAo1raU9AIrxRcyJKVmZmtbSrr3JlUGCMYAAAAo1skzRYdid8ZVVARjAAAAFCuiVpBHx1VUBGMAAAAUq0tsuKLNQSqqKZtJ+d0pusSGe3NaHkcwBgAAQLF8fUyaMSROkgqEY9v3M4bEVfp+xgRjAAAAlGhAm2gtGN1RUWbHcokoc5AWjO5YJfoYs8EHAAAAnDKgTbT6xkWx8x0AAADg62NSfNM65T2NMkEpBQAAACCCMQAAACCJYAwAAABIIhgDAAAAkrj5DgAAoFKyWI0q2x2ivBCMAQAAKpmEpFTNWp2s1Mxs+7Foc5BmDImrEv2EywulFAAAAJVIQlKqJi7d6RCKJSktM1sTl+5UQlJqOc2s8iMYAwAAVBIWq6FZq5NlFHLOdmzW6mRZrIWNQEkIxgAAAJXEtpT0AivFFzIkpWZma1tKuvcmVYUQjAEAACqJk2eKDsXujIMjgjEAAEAlEVEryKPj4IhgDAAAUEl0iQ1XtDlIRTVlMym/O0WX2HBvTqvKIBgDAABUEr4+Js0YEidJBcKx7fsZQ+LoZ+wmgjEAAEAlMqBNtBaM7qgos2O5RJQ5SAtGd6SPcSmwwQcAAEAlM6BNtPrGRbHznYcRjAEAACohXx+T4pvWKe9pVCmUUgAAAAAiGAMAAACSCMYAAACAJIIxAAAAIIlgDAAAAEgiGAMAAACSCMYAAACAJIIxAAAAIIlgDAAAAEgiGAMAAACSCMYAAACAJIIxAAAAIIlgDAAAAEgiGAMAAACSyjkYWywWPfroo4qNjVVwcLCaNm2qxx9/XIZh2McYhqHHHntM0dHRCg4OVp8+fXTw4EGH66Snp2vUqFEKDQ1VWFiYxo8fr7Nnz3r75QAAAKASK9dg/PTTT2vBggV68cUX9d133+npp5/WvHnz9MILL9jHzJs3T88//7wWLlyorVu3qkaNGurfv7+ys7PtY0aNGqX9+/dr3bp1WrNmjb7++mvdeeed5fGSAAAAUEmZjAuXZ71s8ODBioyM1KJFi+zHhg8fruDgYC1dulSGYah+/fp68MEH9dBDD0mSMjMzFRkZqSVLlmjEiBH67rvvFBcXp+3bt6tz586SpISEBA0cOFA///yz6tevX+I8srKyZDablZmZqdDQ0LJ5sQAAAHCbN/Jaua4YX3HFFVq/fr1++OEHSdKePXv0zTff6Nprr5UkpaSkKC0tTX369LE/xmw2q2vXrkpMTJQkJSYmKiwszB6KJalPnz7y8fHR1q1bC33enJwcZWVlOXwBAACgevMrzyd/+OGHlZWVpZYtW8rX11cWi0Vz5szRqFGjJElpaWmSpMjISIfHRUZG2s+lpaUpIiLC4byfn5/Cw8PtYy42d+5czZo1y9MvBwAAAJVYua4Yv/fee1q2bJmWL1+unTt36s0339S//vUvvfnmm2X6vNOnT1dmZqb966effirT5wMAAEDFV64rxlOnTtXDDz+sESNGSJLatm2ro0ePau7cuRozZoyioqIkSSdOnFB0dLT9cSdOnFCHDh0kSVFRUTp58qTDdfPy8pSenm5//MUCAwMVGBhYBq8IAAAAlVW5rhifO3dOPj6OU/D19ZXVapUkxcbGKioqSuvXr7efz8rK0tatWxUfHy9Jio+PV0ZGhnbs2GEfs2HDBlmtVnXt2tULrwIAAABVQbmuGA8ZMkRz5sxRo0aN1Lp1a+3atUvPPvusxo0bJ0kymUy6//779cQTT6hZs2aKjY3Vo48+qvr16+v666+XJLVq1UoDBgzQhAkTtHDhQp0/f16TJ0/WiBEjnOpIAQAAAEjlHIxfeOEFPfroo7rnnnt08uRJ1a9fX3fddZcee+wx+5h//OMf+v3333XnnXcqIyNDV155pRISEhQUFGQfs2zZMk2ePFm9e/eWj4+Phg8frueff748XhIAAAAqqXLtY1xR0McYAACgYqvyfYwBAACAioJgDAAAAIhgDAAAAEgiGAMAAACSCMYAAACAJIIxAAAAIIlgDAAAAEgiGAMAAACSCMYAAACAJIIxAAAAIIlgDAAAAEgiGAMAAACSCMYAAACAJIIxAAAAIEnyK+8JAACAqsViNbQtJV0nz2QrolaQusSGy9fHVN7TAkpEMAYAAB6TkJSqWauTlZqZbT8WbQ7SjCFxGtAmuhxnBpSMUgoAAOARCUmpmrh0p0MolqS0zGxNXLpTCUmp5TQzwDkEYwAAUGoWq6FZq5NlFHLOdmzW6mRZrIWNACoGSikAAIDLLq4jthpGgZXiCxmSUjOztS0lXfFN63hvooALCMYAAJShqngjWmF1xGHB/k499uSZosMzUN4IxgAAlJGqeCOarY744oKIjD/OO/X4iFpBnp8U4CHUGAMAUAaq4o1oxdURl8Sk/H8UdIkN9/S0AI8hGAMA4GFV9Ua0bSnpxdYRF8VWODJjSFylLyNB1UYwBgDAw0oKkBfeiFaZOFsffHG9cZQ5SAtGd6y05SOoPqgxBgDAw5wNkJXtRjRn64NfuqWjfHxMVeqGQ1QPBGMAADzM2QBZ2W5E6xIbrmhzkNIyswstEzEpf3W4W9M6BGFUSpRSAADgYbYAWVQ0rKw3ovn6mDRjSJwkFXht1BGjKiAYAwDgYVU5QA5oE60Fozsqyuy42k0dMaoCk2EYleuW2DKQlZUls9mszMxMhYaGlvd0AAAVlKubdVTFPsY2VXHjElRs3shrBGMRjAEAJXM35BIgAc8gGHsJwRgAUJyidnuzxVtKCICy5428Ro0xAADFqKqbdQAoiGAMAEAxqupmHQAKIhgDAFCMqrpZB4CC2OADAIBiVKbNOrjRDygdgjEAAMVwdre38t6soyq3hgO8hVIKAACKURk267B1zbi4FjotM1sTl+5UQlJqOc0MqFwIxgAAlKAi7/ZG1wzAcyilAADACQPaRKtvXFSFq+F1pWtGfNM63psYUAkRjAEAcJKvj6ncwmVRN9bRNQPwHIIxAAAVXHE31lWmrhlARUeNMQAAFVhJN9b99nuOos1BBW4MtDEpP0SXd9cMoDIgGAMAUEE5c2Pd4598p0cHVeyuGUBlQTAGAKCCcvbGuto1Aips1wygMqHGGACAclTcbnWu3Fg3tEODCtk1A6hMCMYAAJSTknarc/XGuvLsmgFUBZRSAABQDpzZrc62HTU31gHeQTAGAMDLnN2tTlKF344aqEoIxgAAeJkru9VV5O2ogaqGGmMAALzM1d3qKup21EBVQzAGAMDL3NmtjhvrgLJHKQUAAF7GTXVAxUQwBgDAy3x9TNxUB1RABGMAAErJYjW06dAp/euzA/rXZ99r08FTslgL6znxF26qAyoek2EYxf/JrQaysrJkNpuVmZmp0NDQ8p4OAKASSUhK1cMr9ynj3HmH42Eh/npqWNsSA25xO98B+Is38hrBWARjAIDrLFZDL244pPmf/1DsuIWs/gIe4Y28RlcKAABclJCUqpkf71daVk6JY2d+vF9946JYBQYqAWqMAQBwgW0rZ2dCsSSlZeVoW0p6Gc8KgCcQjAEAcFJxWzkXx9kNPQCUr3INxo0bN5bJZCrwNWnSJElSdna2Jk2apDp16qhmzZoaPny4Tpw44XCNY8eOadCgQQoJCVFERISmTp2qvLy88ng5AIAqrqStnIvi7IYeAMpXuQbj7du3KzU11f61bt06SdLf/vY3SdIDDzyg1atXa8WKFfrqq690/PhxDRs2zP54i8WiQYMGKTc3V5s3b9abb76pJUuW6LHHHiuX1wMAqNrcWfmNCg1kow6gkqhQXSnuv/9+rVmzRgcPHlRWVpbq1aun5cuX68Ybb5Qkff/992rVqpUSExPVrVs3ffrppxo8eLCOHz+uyMhISdLChQs1bdo0/frrrwoICHDqeelKAQBwRuLh0xr52haXHkNXCsAzvJHXKkyNcW5urpYuXapx48bJZDJpx44dOn/+vPr06WMf07JlSzVq1EiJiYmSpMTERLVt29YeiiWpf//+ysrK0v79+4t8rpycHGVlZTl8AQBQkpK2cr5QWIg/oRioZCpMMP7oo4+UkZGh22+/XZKUlpamgIAAhYWFOYyLjIxUWlqafcyFodh23nauKHPnzpXZbLZ/NWzY0HMvBABQZRW3lbPNtW0itWx8V+14pC+hGKhkKkwwXrRoka699lrVr1+/zJ9r+vTpyszMtH/99NNPZf6cAICqoaitnKPNQVo4uqMWjO6s7s3q0rcYqIQqxAYfR48e1eeff66VK1faj0VFRSk3N1cZGRkOq8YnTpxQVFSUfcy2bdscrmXrWmEbU5jAwEAFBgZ68BUAAKqTAW2i1Tcuiq2cgSqmQqwYL168WBERERo0aJD9WKdOneTv76/169fbjx04cEDHjh1TfHy8JCk+Pl779u3TyZMn7WPWrVun0NBQxcXFee8FAACqHV8fk+Kb1tHQDg0U37QOoRioAsp9xdhqtWrx4sUaM2aM/Pz+mo7ZbNb48eM1ZcoUhYeHKzQ0VPfee6/i4+PVrVs3SVK/fv0UFxenW2+9VfPmzVNaWpoeeeQRTZo0iRVhAAAAuKTcg/Hnn3+uY8eOady4cQXOzZ8/Xz4+Pho+fLhycnLUv39/vfzyy/bzvr6+WrNmjSZOnKj4+HjVqFFDY8aM0ezZs735EgAAhbBYDUoNAFQqFaqPcXmhjzEAeFZCUqpmrU522CUu2hykGUPivNKpgVAOVD3eyGvlvmIMAKhaEpJSNXHpTl286pKWma2JS3dqQRn39i3vUA6g8qoQN98BAKoGi9XQrNXJBUKxJPuxWauTZbGWzX9W2kL5haFY+iuUJySllsnzAqgaCMYAAI/ZlpJeIJReyJCUmpmtbSnpHn3e3DyrXvv6R015d0+5hXIAlR+lFAAAjzl5puhQ7M44Z8xdm6zXNqaopLx7YSiPb1rHY88PoOogGAMAPCaiVlDJg1wYV5K5a5P1ytcpLj3Gk6EcQNVCMAYAeEyX2HBFm4OUlpldaEmDSVKUOb9LhDsu7DYRHhyg1za6Foolz4VyAFUPwRgA4DG+PibNGBKniUt3yiQ5hGNbs7QZQ+Lcap1WWLcJV5Q2lAOo+rj5DgDgUQPaRGvB6I6KMjuuzEaZg9xu1VZUtwlnlTaUA6geWDEGAHjcgDbR6hsX5ZFNNoprAeesKPoYA3ACwRgAUCZ8fUwe6f5QUgu44pgk/XdcF11xaV1WigGUiGAMAChUeWyrXNhzlqaLxJ09Y9WjeT0PzhBAVUYwBgAUUB7bKhf1nCMub+TytXxM0oQesZo+MM6TUwRQxZkMw6j2WwBlZWXJbDYrMzNToaGh5T0dAChXthvdLv7hYFsrdvcGutI8pznEX5nnzhfZAi6iVoDu6NFUP/12TjHhIbo1vrEC/Li/HKhKvJHXWDEGANgVd6ObofwQOmt1svrGRXmsrMKZ5zRd8OvCWsDNGtqGG+sAlBr/nAYA2JV0o9uF2yp78zl/O3deD/Rp5tEWcABwMVaMAQB2zt7o5sltlZ29VuO6NfTNtF5evyEQQPVBMAYA2Dm7XbInt1V25Tk91QIOAApDKQUAVEMWq6HEw6e1avcvSjx8WhZrfuVul9hwRZuDVNQarEn5nSI8ua1yeTwnABSGFWMAqGZKasU2Y0icJi7dWeSNbp7eVtnXx+T15wSAwrBiDADViK0t2sU3u6VlZmvi0p1KSErVgDbRWjC6o1dvdCuP5wSAi9HHWPQxBlA9WKyGrnx6Q5EdIEzKD6LfTOslXx9Thdn5jpViABJ9jAEAHuRKK7b4pnXK5UY3bq4DUJ4opQCAaqI8WrEBQGVCMAaAaqI8WrEBQGVCKQUAVBO2tmhpmdmFbr9sqzEuqi0a9b8AqjqCMQBUE6Vpi1ZSizcAqAoopQCAasTVtmgWq6HnPv9Bd5fQ4g0AqgLatYl2bQCqH2fKIhKSUjXz42SlZRV9M97FLd4AoKzQrg0AUCZKaotm2wikpJWTi1u8AUBlRikFAMCBxWpo1urkEkPxhWjxBqAqIBgDAByUtBFIYWjxBqAqoJQCACqZsm6b5srqb0kt3gCgMiEYA0Al4o22aa6u/hbV4g0AKhtKKQCgkrDdEFfWbdNsG4GUFHWjQgMLbfEGAJUVwRgAKoHiboizHZu1OlkWa+k7cNo2ApFUZDh+oE9zbXq4N6EYQJVCMAaASqCkG+IubJvmCUVtBBJtDtLC0R11X59mlE8AqHKoMQaASsDZG+I82TZtQJto9Y2LKtMb/QCgIiEYA0Al4OwNcZ5um1bSRiAAUJVQSgEAlUBJN8SZlF/mQNs0AHAfwRgAKhCL1VDi4dNatfsXJR4+bb+Zrrgb4mzf0zYNAEqHUgoAqCBK6lFsuyHu4jFRHu5jDADVlckwjNL39qnksrKyZDablZmZqdDQ0PKeDoBqyNaj+OK/kG3rvxf2Cy7rne8AoCLyRl5jxRgAyllJPYpNyu9R3DcuSr4+Jm6IA4AyQo0xAJQzb/coBgAUjhVjAChnzvYe/mTfcUmidAIAyojTwdhiscjX17cs5wIA1ZKzvYeXbjmmpVuOOdyQBwDwHKdLKRo0aKCHH35YP/zwQ1nOBwCqpKLasEkl9yi+WGpmtiYu3amEpNSymSwAVFNOB+NJkybp/fffV6tWrdSjRw8tWbJE586dK8u5AUCVkJCUqu5PbdDI17bovnd2a+RrW9T9qQ32YHthj2JnGcq/Ie/CgA0AKB2ng/Gjjz6qQ4cOaf369WrSpIkmT56s6OhoTZgwQVu3bi3LOQJApbV2b6ruXrpTaVmOdcRpWdm6+6JVX3OIv0vX5oY8APAsl7tSXH311XrzzTeVlpamf//73/ruu+8UHx+v1q1b69lnny2LOQJApbR273FNWr6z2DHTV+7T/HUHdPfSnco4d97l53D2xj0AQMk8ssHHJ598ottuu00ZGRmyWCyemJdXscEHAE9LSMpfKS5rb0/oRk9jANWCN/Ka232Mz507pyVLluiqq67Sddddpzp16mjOnDmenBsAVEoWq6GZH+8v8+eJNufvegcA8AyX+xhv3rxZb7zxhlasWKG8vDzdeOONevzxx9WzZ8+ymB8AVDrbUtKVlpVTps9hkjRjSBz9jAHAg5wOxvPmzdPixYv1ww8/qHPnznrmmWc0cuRI1apVS1J+n+MTJ06ofv36ZTZZAKgMyrrulz7GAFA2nA7GzzzzjEaPHq0VK1aoTZs2Bc4nJSWpY8eOlbLGGAA8ydkNO5xlDvbT5GuaqW7NAEWZg9n5DgDKiNPB+Pjx4/L3d62VEABUdRaroW0p6Tp5JlsRtfJrfjvF1FbtEH/95kaXiYuZJD09vB2rwwDgBU4HY0IxADhKSErVrNXJSs38q3SiZqCfrIahc7ml/9+zqNBAzbyuNaEYALzE5ZvvAAD5oXji0p26uN/l2Zw8j1z/gT7NNbnXpZRMAIAXOR2M9+7dW+z5AwcOlHoyAFAZWKyGZq1OLhCKPYEb6wCg/DgdjDt06CCTyaTC9gOxHTeZXF/Z+OWXXzRt2jR9+umnOnfunC699FItXrxYnTt3liQZhqEZM2botddeU0ZGhrp3764FCxaoWbNm9mukp6fr3nvv1erVq+Xj46Phw4frueeeU82aNV2eDwAUxVZPvOnQrw7lE54wvntj9YmL4sY6AChHTgfjlJQUjz/5b7/9pu7du+uaa67Rp59+qnr16ungwYOqXbu2fcy8efP0/PPP680331RsbKweffRR9e/fX8nJyQoKyr/ze9SoUUpNTdW6det0/vx5jR07VnfeeaeWL1/u8TkDqJ4Kqyf2hDo1AjTnhjasEANABeCRLaHd9fDDD2vTpk3auHFjoecNw1D9+vX14IMP6qGHHpIkZWZmKjIyUkuWLNGIESP03XffKS4uTtu3b7evMickJGjgwIH6+eefneqrzJbQAIqzdm+q7lnu+e2dw2v4a8v0Pgrwc3sTUgCoNirUltAHDx7UyJEjlZWVVeBcZmambrnlFv34448uPfnHH3+szp07629/+5siIiJ02WWX6bXXXrOfT0lJUVpamvr06WM/Zjab1bVrVyUmJkqSEhMTFRYWZg/FktSnTx/5+Pho69athT5vTk6OsrKyHL4AoDBr9x7X5Lc9G4pNf349eUNbQjEAVCBO/438zDPPqGHDhoUmdLPZrIYNG+qZZ55x6cl//PFHe73wZ599pokTJ+rvf/+73nzzTUlSWlqaJCkyMtLhcZGRkfZzaWlpioiIcDjv5+en8PBw+5iLzZ07V2az2f7VsGFDl+YNoHpISErVPct3yerh/1eLMgdpweiOlE8AQAXjdI3xV199paVLlxZ5/qabbtItt9zi0pNbrVZ17txZTz75pCTpsssuU1JSkhYuXKgxY8a4dC1XTJ8+XVOmTLF/n5WVRTgGICn/Brsth09r84+n9PpG1/4XrCgmk/T3Xs3UpF4N+yYg3GAHABWP08H42LFjBVZmL1S3bl399NNPLj15dHS04uLiHI61atVKH3zwgSQpKipKknTixAlFR/+1snLixAl16NDBPubkyZMO18jLy1N6err98RcLDAxUYGCgS3MFUPUlJKXq4ZX7lOGBHesu9NLIjhrYjtVhAKjonC6lMJvNOnz4cJHnDx065HIhdPfu3Qv0P/7hhx8UExMjSYqNjVVUVJTWr19vP5+VlaWtW7cqPj5ekhQfH6+MjAzt2LHDPmbDhg2yWq3q2rWrS/MBUH2t3Xtcdy/d6dFQXDvEXwtHE4oBoLJwesW4Z8+eeuGFF9SrV69Czz///PPq0aOHS0/+wAMP6IorrtCTTz6pm266Sdu2bdOrr76qV199VVJ+f+T7779fTzzxhJo1a2Zv11a/fn1df/31kvJXmAcMGKAJEyZo4cKFOn/+vCZPnqwRI0Y41ZECANbuTdXkt3e5/fiwEH+HQB0W7K+x3Rtrcq9mlEwAQCXidLu2Xbt2KT4+XoMHD9Y//vEPtWjRQpL0/fffa968efrkk0+0efNmdezY0aUJrFmzRtOnT9fBgwcVGxurKVOmaMKECfbztg0+Xn31VWVkZOjKK6/Uyy+/rObNm9vHpKena/LkyQ4bfDz//PNOb/BBuzag+rFt1rEuOU1vbDri9nXu791M9/Zupm0p6Tp5JpsaYgAoI97Iay71MV6zZo3GjRun06dPOxyvU6eOXn/9dV133XUen6A3EIyB6mXt3uN6ZFWS0n8vXdlEWLCfdjzajxAMAF7gjbzmdCmFJA0ePFhHjx5VQkKCDh06JMMw1Lx5c/Xr108hISFlMkEA8JTcPKvGLNqqxJR0j1zvqeHtCMUAUIW4FIwlKTg4WDfccENZzAUAyszctcl65WvPbG0fbQ7SjCFx9CEGgCrG5WAMAJWNp0LxuO6N1TcuihpiAKiiCMYAqhTbTXW2G+E6NAzTaxtLF4rDgv301PB2rBADQBVHMAZQZSQkpWrW6mSlZmbbj9UK8i3Vls62rhOsEANA1UcwBlAlJCSlauLSnbo4A5/Jtrh1PeqIAaD6cSoYZ2VlOX1B2p0B8DaL1dCs1ckFQrE7xl4Ro36to6kjBoBqyKlgHBYWJpPJuR8QFot7qzMA4K5tKekO5RPuuqtnrKYPjPPAjAAAlZFTwfiLL76w//rIkSN6+OGHdfvttys+Pl6SlJiYqDfffFNz584tm1kCQDFOnildKA7y99Gzf+ugge0omwCA6sypYHzVVVfZfz179mw9++yzGjlypP3Yddddp7Zt2+rVV1/VmDFjPD9LAC6xWA1t+fG0Eg+flmQovklddWtap0qUBvyRa9GTa5N15PQ5Na4Ton8OjFNErSCnHmsySRfu9WmSNLhdlP4zomOVeG8AAKXj0pbQkhQSEqI9e/aoWbNmDsd/+OEHdejQQefOnfPoBL2BLaFRlSQkperhlfuUcc5xu+OwEH89Naxtpb2ZLDfPqsHPb9QPJ88WONenVYT2H89SWmZ2oXXGJklR5iBtePBqLd96VEfTzykmPES3xjdWgJ9Pmc8dAFB63shrLv9EaNiwoV577bUCx19//XU1bNjQI5MC4J6EpFTdvXRngVAsSRnnzuvupTuVkJRaDjNzn8Vq6N7lO9T8kU8LDcWS9Pl3J1WvVoCk/BB8Idv3M4bEKTjAV+N7NNHsoW00vkcTQjEAwIHL7drmz5+v4cOH69NPP1XXrl0lSdu2bdPBgwf1wQcfeHyCAJyTm2fV9JX7Shw38+P96hsXVSlKB9buTdWUFbuVfd5a4ti9P2fp+Zvaa+5nBxxuxIui7RoAwEkuB+OBAwfqhx9+0IIFC/T9999LkoYMGaK7776bFWOgnCQkpeqfHybpt0JWii+WlpWjbSnpim9axwszc5872zhvP/abvpnWy2HnO9quAQCc5dYGHw0bNtSTTz7p6bkAcENRG1sUp7RdHMra2r3HXQ7FknTk9Dn5+pgqfOgHAFRMbhXYbdy4UaNHj9YVV1yhX375RZL01ltv6ZtvvvHo5AAUz92NLZzt4lAeLFZDj6xKcuuxrAsDAErD5WD8wQcfqH///goODtbOnTuVk5MjScrMzGQVGfAydza2iAoNVJfY8DKaUeltS0lX+u8ll4QUJvmX32SxemL/OwBAdeRyMH7iiSe0cOFCvfbaa/L397cf7969u3bu3OnRyQEonjslETOva12ha25LU+Zx6pxF21LSPTgbAEB14nIwPnDggHr27FnguNlsVkZGhifmBMBJrpREhIX4a+HojhW+O0Npyzwqev00AKDicvnmu6ioKB06dEiNGzd2OP7NN9+oSZMmnpoXgCJYrIa960LdGoGKCg3SiazCN7aQpJqBvlpwSydd0axuhV4ptukSG65oc5DLJSI2Fbl+GgBQsbkcjCdMmKD77rtPb7zxhkwmk44fP67ExEQ99NBDevTRR8tijgD+lJCUqlmrkx1CY1iIvwzl33h2YTi2ReB//a29erSo58VZus8W+ge2idKiTUdcfnxkrYAKXT8NAKjYXA7GDz/8sKxWq3r37q1z586pZ8+eCgwM1EMPPaR77723LOYIQPktzO5ZvqvA8cw/exebQ/wddryrbBtbFBb6fUySK/fSzRraplKsigMAKiaTYRhu3cKdm5urQ4cO6ezZs4qLi1PNmjU9PTev8cbe20BprN2bqslv7ywyJJokRYYG6t83ddCpszmVbmOLonox21bBe7Wspw3f/1rk4wP8fPT8iA6V5h8BAADXeSOvuXzz3bhx43TmzBkFBAQoLi5OXbp0Uc2aNfX7779r3LhxZTFHoFpbuzdV9ywvOhRL+eExLStHPiaThnZooPimdSpFKLZYDW06dEoPf7Cv0BppW4nId6ln9PItHRVtdqwfDvb30d+vuVTfzR5AKAYAlJrLK8a+vr5KTU1VRESEw/FTp04pKipKeXl5Hp2gN7BijIpq7d7jmvz2LqfLCZ4b0UFDOzQo20l5SGGlE8V5e0I3dYkNZ7tnAKimvJHXnK4xzsrKkmEYMgxDZ86cUVDQXys3FotFa9euLRCWAbgvISm10Jri4lSWjgzubmPNds8AgLLkdDAOCwuTyWSSyWRS8+bNC5w3mUyaNWuWRycHVFe2rZ5dEW0OqhQdGariNtYAgKrB6WD8xRdfyDAM9erVSx988IHCw//6ARwQEKCYmBjVr1+/TCYJVCcWq6Elm1Jc7uM7Y0hchS0ruLD38qkzOS69NpPyO2xUhtAPAKjcnA7GV111lSQpJSVFjRo1kslUMX8AA5WZq3W3Un5LsxdHXlZhbz5z5zXZ2P6WqcihHwBQdbjcx3jDhg2qWbOm/va3vzkcX7Fihc6dO6cxY8Z4bHJAdeJO3a0kvTiyowa2q7ih2J3XZFPZejEDACo3l9u1zZ07V3Xr1i1wPCIiQk8++aRHJgVUN+7U3fqYpJdvqbihODfPqn9+WHgbtpKEhfhr2fiu+mZaL0IxAMBrXF4xPnbsmGJjYwscj4mJ0bFjxzwyKaC62XL4tMulBi+OvKzChuKEpFT988Mkpf9+vuTBF7AVSzw1rK26Nyv4D3AAAMqSyyvGERER2rt3b4Hje/bsUZ06tFECXJWQlKpJy3c6PT7aHKSFoztqYLuKebOrrXwi/fdclx8bZQ7SgtEdWSUGAJQLl1eMR44cqb///e+qVauWevbsKUn66quvdN9992nEiBEenyBQlblag/vooFa6vXtsmdyIdmHnCHc3z3CnJOTRQa1Ut1YgG3YAAMqdy8H48ccf15EjR9S7d2/5+eU/3Gq16rbbbqPGGHCBKyHS1rKsrEJxYZ0jot248W1bSrrTJSFl/ZoAAHCVy8E4ICBA7777rh5//HHt2bNHwcHBatu2rWJiYspifkCVtfnQKZfqisuqZdnavam6p5BSjrTMbE1curPE0oYLV5oPnjjr0nPThg0AUJG4HIxtmjdvXugOeABKNndtsl79OsWpsWHB/npqeNsyqbtdu/e4Jr9d+LbThvJXdWetTlbfuKgCAdZiNfTihoNavOmIMv5w7Sa78Br+evKGsnlNAAC4y6lgPGXKFD3++OOqUaOGpkyZUuzYZ5991iMTA6qquWuT9YqToViSXhrVUd0v9XyHhoSkVN2zvPBQbGNISs3M1raUdMU3/evm2oSkVD28cp8yzrkWiCWpTo0AJU7vrQA/l+/9BQCgTDkVjHft2qXz58/bf10UdsMDivdHrkWvbnQuFNtqcLs18Xy3F1t9s7NOnvmr5MPdTTtsfzvMuaENoRgAUCE5FYy/+OKLQn8NwHlr9x7Xgyv2yHAhUZZVDa4rN8lJUkStIEnudZ2wYRc7AEBF53aNMQDnuVo+ERLgq2dval+qEFlc+7ULV4BLEm3Of6zkeqCefE1TNYusRSs2AECl4FQwHjZsmNMXXLlypduTAaqitXtTXQrFkvRAn2Zuh+Kiboq7sP2abQXYGSMub6Q1e48rolaQ0rJc252v+6X1HGqTAQCoyJwKxmaz2f5rwzD04Ycfymw2q3PnzpKkHTt2KCMjw6UADVQHFquhR1YlufQYH5M05oqC2647Y83u43rogz3KPm8tcO7C9mt946IUbQ5SWmZ2kWURJpNkDvLT/M9/sB8Lr+Hv9FwuXGkGAKAycCoYL1682P7radOm6aabbtLChQvl6+srSbJYLLrnnnsUGhpaNrMEKqltKekub408oUesWzenTfjvdq1LPlnk+Yvbr80YEqeJS3fK9Oe5AuMNKeOPPIdj6b8734WCHsUAgMrG5Z++b7zxhh566CF7KJYkX19fTZkyRW+88YZHJwdUdv/bn+r0WJNJuqtnrKYPjHNqvMVqKPHwaa3a/YsmLdtRbCi2ubD92oA20VowuqOizI5lFc5m2aKGhYX4a2EJm4IAAFARuXzzXV5enr7//nu1aNHC4fj3338vq7Xgf98C1dWcT5K1ePNRp8aG+Ptox6P9FBzgW/JgFb6FsytsN98NaBOtvnFR2paSrnXJaXpj0xFZnWw5UbtGgMNqeFiIv8ZeEavJvS5lpRgAUCm5HIzHjh2r8ePH6/Dhw+rSpYskaevWrXrqqac0duxYj08QqIyeWJ2s1zc5f8Pdv/7W3qVQ7E4f4QtdePOdr49JXWLDNeW93S5d49FBrRRlDi606wUAAJWRy8H4X//6l6KiovTvf/9bqan5/00cHR2tqVOn6sEHH/T4BIHK5vE1+7Vo0xGnx9/VM1YD29V3amxp+gjbFHZTnKtt2CQpyhxMxwkAQJXicjD28fHRP/7xD/3jH/9QVlaWJHHTHfCnOZ8ka9E3R5weP/aKmBJrii/sR3zqTI7b5RM2M4bkP1/i4dP21V5X2rDZduSj4wQAoKpxa4OPvLw8ffnllzp8+LBuueUWSdLx48cVGhqqmjVrenSCQGWxdu9xvebkds82/VoXf4NaaWuJL2QO8tXTN7aXJF359AaHa7rShk2i4wQAoGpyORgfPXpUAwYM0LFjx5STk6O+ffuqVq1aevrpp5WTk6OFCxeWxTyBCs1iNfTQij0uPSa8hn+Rq675m3QccughXBqdYsL03l1XaF1yWqH1yc62YYsKDdTM61rTcQIAUCW5HIzvu+8+de7cWXv27FGdOn/VF95www2aMGGCRycHVBY3vbJZ5wrZVKM4Twxt47DqaiuZ+Dw5TSt3/azfzuUV82jnmEzSHVfG6v8GxSk3z6p/frivxPrkovoaP9CnOR0nAABVmsvBeOPGjdq8ebMCAgIcjjdu3Fi//PKLxyYGVBZzPknWjqMZLj1mQg/HG+48UTJhC7Q3drxEIYG+igkP0S1dY7T7pwzNXr1fK3b8rDPZJYfti9uwXbiVNAAAVZnLwdhqtcpisRQ4/vPPP6tWrVoemRRQWeTmWfX6N67VFd/RPX8F18YT7dek/BviLgywCUmp6vXvL10O27RhAwBUVy4H4379+uk///mPXn31VUmSyWTS2bNnNWPGDA0cONDjEwQqsrcSj8hwIdGOv7KxHhn8Vygubfu1Rwe1Ut1agQUCbGnCNm3YAADVlVt9jAcMGKC4uDhlZ2frlltu0cGDB1W3bl29/fbbZTFHoMI6mn7O6bGD2kbp0cGtPdJ+zdYy7fbusQVWc90N27RhAwBUdy4H44YNG2rPnj169913tWfPHp09e1bjx4/XqFGjFBwcXBZzBCocW7jN+sO5bg5B/j56fmRHj7ZfK6plmjubdZR0TQAAqgOTYTj/H8Hnz59Xy5YttWbNGrVq1aos5+VVWVlZMpvNyszMZLMSlMidcDv/pvbadOi03t/5c6mfv7Cb4f7ItejJtck6cvqcrFZDmw6fduma4TX89eQNbbnBDgBQYXkjr7m0Yuzv76/s7NKvdAGV1dq9qbpn+U6XHhNTJ1gPrtgja2nvrpP0QJ9mmtyrmcOq7oT/bte65JNuX7NOjQAlTu+tAD+f0k8QAIBKzOWfhJMmTdLTTz+tvLzS91idOXOmTCaTw1fLli3t57OzszVp0iTVqVNHNWvW1PDhw3XixAmHaxw7dkyDBg1SSEiIIiIiNHXqVI/MDbjY2r3HNflt50OxySS1axCqo6f/KHUojjYHaeHojrqvT3OHUHzHm9vcDsWmP7/m3NCGUAwAgNyoMd6+fbvWr1+v//3vf2rbtq1q1KjhcH7lypUuXa9169b6/PPP/5qQ319TeuCBB/TJJ59oxYoVMpvNmjx5soYNG6ZNmzZJkiwWiwYNGqSoqCht3rxZqampuu222+Tv768nn3zS1ZcGFCkhKVX3LN/l1Ni+rSIUZQ5S2wZmTVu5z+3nrFMjQEM71FffuKhCW6at2vWLPv/uV7evf3F7NwAAqjuXg3FYWJiGDx/uuQn4+SkqKqrA8czMTC1atEjLly9Xr169JEmLFy9Wq1attGXLFnXr1k3/+9//lJycrM8//1yRkZHq0KGDHn/8cU2bNk0zZ84ssAmJTU5OjnJycuzfZ2Vleez1oOqxdXlw1vYjvynDyZvyLlZU+7WLJSSl6r53d7t8/fAa/rqhQwP1KSJsAwBQnbkcjBcvXuzRCRw8eFD169dXUFCQ4uPjNXfuXDVq1Eg7duzQ+fPn1adPH/vYli1bqlGjRkpMTFS3bt2UmJiotm3bKjIy0j6mf//+mjhxovbv36/LLrus0OecO3euZs2a5dHXgarL1S4P7oTi4tqvXchiNbTl8Gk9/IHzK9FtG4Tqjh5N2KwDAIASOF1YaLVa9fTTT6t79+66/PLL9fDDD+uPP/4o1ZN37dpVS5YsUUJCghYsWKCUlBT16NFDZ86cUVpamgICAhQWFubwmMjISKWlpUmS0tLSHEKx7bztXFGmT5+uzMxM+9dPP/1UqteBqu3kGe/ccFpSq7SEpFRd+fQGjVq01aXw3aFhmIZ2aKD4pnUIxQAAFMPpFeM5c+Zo5syZ6tOnj4KDg/Xcc8/p5MmTeuONN9x+8muvvdb+63bt2qlr166KiYnRe++9V6Y9kQMDAxUYGFhm10fVElErqEyvHxUaqJnXtS6y1tdiNfTihkOa//kPbl3/nwPjSh4EAACcXzH+73//q5dfflmfffaZPvroI61evVrLli2T1Wr12GTCwsLUvHlzHTp0SFFRUcrNzVVGRobDmBMnTthrkqOiogp0qbB9X1jdMuCOLrHhijaXTTge3C5amx7uXWQoXrs3VZ2fWOd2KO4bF6HgAN/STBEAgGrD6WB87NgxDRw40P59nz59ZDKZdPz4cY9N5uzZszp8+LCio6PVqVMn+fv7a/369fbzBw4c0LFjxxQfHy9Jio+P1759+3Ty5F/tqtatW6fQ0FDFxbFKBs/w9THpuvae7dzgY5Lu6hmrF2/pWGR5w9y1ybpn+U79ds69G/n6tKqn1267vDTTBACgWnG6lCIvL09BQY6rZv7+/jp/3r0f2pL00EMPaciQIYqJidHx48c1Y8YM+fr6auTIkTKbzRo/frymTJmi8PBwhYaG6t5771V8fLy6desmSerXr5/i4uJ06623at68eUpLS9MjjzyiSZMmUSoBt9hubkv88ZQkk+Kb1tHljcP18Z7UUl87NMhPQzvUV+M6NXRrfONiewev3Xtcr3yd4vZzPXdzBw29rIHbjwcAoDpyOhgbhqHbb7/dIXBmZ2fr7rvvduhl7Eof459//lkjR47U6dOnVa9ePV155ZXasmWL6tWrJ0maP3++fHx8NHz4cOXk5Kh///56+eWX7Y/39fXVmjVrNHHiRMXHx6tGjRoaM2aMZs+e7fQcAJuEpFQ9vHKfMi5YoX3xi0OqGeinsznubxpjWw+ed2M7p3oGW6yGHlmV5NZzFbZdNAAAcI7JMAyn9uQaO3asUxf0dDs3b/DG3tuo2BKSUnX3Ute2enaWM2HVYjW0+eApfbDrZ/382zl9ezTDpecIC/bXS6M6qlsTOk8AAKomb+Q1p1eMK2PgBZxhsRqa+fF+j193fPfGTm2kkZCUqinv7dG5XItbz2OS9NTwtup+aV03ZwoAACQ3NvhA6VishralpOvkmWw2XKggtqWkKy0rp+SBTnKlnKG0K9WUTgAA4DkEYy9KSErVrNXJDruoEWzKx4X/QDmQdsbt65gkGZIe6NNMjevWcOkfOxaroRlu1hJL0v29m+ne3s34hxUAAB5CMPaShKRUTVy6UxcXdKdlZmvi0p1aMLoj4dhLCvsHirPCa/gr/fe/bs6LcvMfNharoSWbUnTiTK7Lc5DyW73d37e5W48FAACFIxh7gcVqaNbq5AKhWMpfbTRJmrU6WX3jolj9K0P5O8gd1PzPD7r1+GhzkL6aeo12HP2tVKUwpQ3mTwxto4Ht6rv8WAAAUDyCsRdsS0kvNgQZklIzs7UtJV3xTet4b2LVSEJSqmZ+nKy0LNfDqM2MIXEK8PNx6/fIVrrxeXKaFm064vLjr+9QXzdf3oiadAAAyhDB2AtOnnEujDk7Dq4pqozFWbVD/DV3WFu3S11Ks0Is5e+SN+/G9sVuCAIAAEqPYOwFEbWCSh7kwjg4r7gyFmdMuqappvRt4fYqbWlDuSRN6BFLKAYAwAv4aesFXWLDFW0OUlHRyqT8+tUuseHenFa1UFIZS0muvLSe26G4tKHcZMq/yW76wDg3rwAAAFzBirEX+PqYNGNInCYu3Wlv72Vji1wzhsRRO1oG3C1PMSm/40Rp/rHibijvHBOm/q2jNeaKxqwUAwDgRfzU9ZIBbaK1YHRHRZkdyyWizEG0aitD7pSneOofK66Gctv/HLx71xWa0LMJoRgAAC9jxdiLBrSJVt+4KHa+8yJbGUtaZrbTJQ3u9ia+mCuhnP85AACg/BGMvczXx0RLNg9wZWvtEZc3LLZ38X29m6lLbLhOnc0p1T9WLp5Tp5jaTodyT4VxAADgPoIxKh1nt9YuqU2ap7bjtm0csnjTEWX88deueNHmIF3XPlqvfp1SaG25IWlc98bqGxfF/xwAAFABmAzDKE0nqSohKytLZrNZmZmZCg0NLe/poBC21dh1yWl6o5ANMmyR0lavvXZvqu5ZvrPI6z3Qp7km97q0VGE0PxAf0itfH9a5XEuRc7qzZ6w+3pNaYpAHAABF80ZeY8UYFZ4zG2RcuLV2Xp6he9/ZVeRYk6R3th/T5F6XlmpOD6/cp4xz54scY5vTx3tSPbKVNAAAKFsEY1RormyQYdtae3IxofjCce5uwZ2QlKq7lxa9Gl3Yc+04+hu15QAAVHAEY1Q4uXlWvZV4REdOn9NHu38u1a5xxXGnx7Ft0w5vPBcAAPAugjEqlLlrk/XaxhRZvVD57mqPY4vV0JJNKW5t2sF23wAAVHwEY1QYc9cm65WvU7zyXHVqBLi0q50zdc5FYbtvAAAqB4IxKoTcPKte2+idUCxJQzvUd/rmN1fqnAvDph0AAFQOBGMvs9XPHk0/p5jwEN0a35itfyW9lXjEK+UTNn3jopwaZ6spdmdqYSH+empYW1qyAQBQSRCMvaiw+tk5a7/ThB6xmj4wrvwm5kVF7Vh3NP2cW9cbe0WMVuz4RWdz8px+TFRooNOlDdtS0l0un6gR4Ks7ezYtdZ9kAADgXQRjLymqftZqyH68qofj4nasiwkPcela0eYgPTqolX7JyHYpFEvSzOtaOx1YXekmERbsr7HdG2tyr2YEYgAAKiGCsRc4Uz/72sYUPdivZZUrq7CtEH+enKZFhexYl5aZrYlLd+qFER3kY1Kx5RQ+JumZ4e1Uv3aITp3N0SOr9iv991yn5xIS4Ktnb2rvUmmDs90kHh3USrd3jyUQAwBQiRGMvcCZ+lmrkT9ufI8m3pmUF7iyY92cT7/X+Ctji/0HxIQesRreuaFb3SsujwnTO3ddUWhwLaq8Q5K6xIYr2hyktMzsQuuMTZKizEGEYgAAqgCCsRc4Wz/rbp1tReTOjnW9WkbKx6QCddg+JtnrsNfuPe5WS7f7+7YoNLgWV94xoE20fH1MmjEkThOX7pTpz7na2K5G1wkAAKoGgrEXOFs/62qdbUXlbieHk2eyNX1gnB7s17LQzh0Wq6FHViW5PJ/aIf7q1qTgdsxFhXdbeceC0R01oE20BrSJ1oLRHQsE6KgLAjQAAKj8CMZecGt8Y81Z+12J9bO3xjf22pzKkjudHKS/6nkD/HwKLSnZcvi00n8/7/J15w5rW2BFt7jwbivvmLU6WX3jouTrY9KANtHqGxdVZMkFAACo/AjGXhDg56MJPWKLLQGY0CO20t14V1RtriudHKS/6nSLa6G2du9xTX1/r0vXjS5mRbek8G4r79iWkq74pvmrzb4+JvuvAQBA1UMw9hJbK7bi6mcrC4vV0AvrD+r1b37U2RyL/bgtiDrbyUFyrk53zifJLu2KF+zvo9dvu1zdmtYp8prOhndXQz4AAKi8CMZeNH1gnO7v00JPrk3WkdPn1LhOiP45ME7BAb7lPTWnWKyGXtxwUC99eVi5edYC51P/rM196ZaOxXZyuFBJdbqPr0nWom9cu9luRJdG6t6sbrFjnA3vroR8AABQuRGMvSghKVUzP05WWlb+KuTGg9K65JOaeV3Fv4Fr7d7jmvrBXv1+wQpxYQxJj3+SrEcHtdKk5bsK7eRgSBrXvbH6xkUVW6c755P9WvTNEZfn2s+J7Z6dbcPm7A55AACg8qtcRa2VWEJSqu5eutMeim3SsrJ199KdSkhKLaeZFc9iNTR5+U7ds3xXiaHYJjUzW7VrBGrB6I6KMjuuuEaZg7RwdEc9NqS14ospdVi797he23jE5flGXxBmLVZDiYdPa9XuX5R4+LQsF9Sw2NqwSX+Vc9jQhg0AgOqJFWMvsFgNPbxyX7Fjpq/cZ++AUFHk3/C2R7/nFiybKMnJM9ka2qGBW50cLFZDUz9w7UY7KT/Q2sLuc58f1OJNKcr4468uFhffjEcbNgAAcCGCsRds+fG0Ms4V32bst3PnteXH0+p+afG1sd7i6g1vF7PV5rrTyeHFDYecXp22iQoN1MzrWkuSOj3+P2X8kVdgzMX9iSXRhg0AANhRSuEFiYdPe3RcWSttKK5TI8Dt2tzcPKte/vKQS4+5r3czbXq4tyTp7qU7Cw3F0l+1zrNWJxcoq4hvWkdDOzQotrwDAABUbQRjr3B2DzhX94rzvPzaXvdDsSQ9PrSNW+EyISlVbWd+ppxCOl4UZUKPWD3Qt7kklViuIjn2JwYAALgQwdgL4ps4Vx7h7Liy4u6Wyxe6q2esBrZzvTbXdnOiK6F4/JWx+r9B+TXFWw6XXK5yIfoTAwCAi1Fj7AWXx4YXaFt2MdOf48rTtpR0t7ZclqSagb6aN7ydBrar7/Jjc/OsevC93S49ZlDbKD06+K9NURJ/POXS4+lPDAAALkYw9oIdR38rsUjC+HNcWW85XNQ2zpJ7q6iXNQzTQ/1bqFsT92pzE5JS9eB7rnW+CPYzqUPD2npsVZJiwkN0a3xjFWy6VrRo+hMDAIBCEIy9oCJsP2zbtW7xpiNFtjBzZRW1RoCvnrmxvVtlEza28glXZecZmrP2O/v3c9Z+p4FtSt7Uw4b+xAAAoDAEYy+oWyPQo+NctXZvqv7xwV6dzSm+hVnfuChFm4McevoWZnDbKD03smOpwqUzvZ2LcvHqu9WQ1uxLU6CfT7E1yiaT9NLIjvQnBgAAheLmOy+wGs51m3B2nLP+2rVuZ6GhWHJsYSblr6YWF3cn9IjVi6M6lXrF1dWb5ZyRaym+HOOlkZeVaoUbAABUbQRjL9ia4lx/YmfHFcdiNbTxwK+6eeFmNf+/tVqzt+Stpi9sYWbbDS76oq2cw2v46+VbLrN3gSgtV2+Wc4ZhSDd2vERRoY5zj/5zG2p3bgwEAADVB6UUXmB1ciHY2XGFya8hPqSXvjykXBdanl3IVuPsnd3gyqbGNyTQV5se7sVOdgAAwGUEYy+oHRLg0XEXslgNPb/+By386keXegAX5sKb79zZytkZtq4YRhltZhITHlJmcwcAAFUbwdgL6tZy8uY7J8fZJCSl6oF3d+uP86ULxJJ3WpglJKVq1urkEm/uc5ePSX+2bgMAAHAdwdgLIpwMvM6Os7Vem//5wdJMy0FZtzBLSErVxKU7y3TT6wk9YhXgR9k8AABwD8HYC6xOFg87M271nuP654f7dCa78C4TrgoL8ddTw9qWaQszi9XQzI+TyywU+5jyQ/H0gZ65MRAAAFRPBGMv2JqS7vS4Hs3rFXrubHaerv7XFzp1Ntdj8xrcLlrPjbiszG9Me3HDQaVleb58ItDPpIf6tdSYKxqzUgwAAEqNYOwVzq6VFhz3R65FV8xdr9/+8FzP35qBvpo3vJ1X2pclJKV6tOTjQvNvoi8xAADwHIKxF8Q3qasXvzjs1LgL3fHmdn3+3UmPzSPA16RJ11yqyb2aeaV9mcVq2DcO8bS7esYSigEAgEcRjL2gW9M6Cgnw1blcS5FjQgJ81e3PFmMWq6G+87/Uj7+e89gcOsWE6b27rvBqP99tKeke70ARGuSnp4a1ZbMOAADgcQRjLykpjpr+HLB2b6qmvr9bv+eWvgWblF828dQNbTW4QwOPXM8VaZl/ePR64TX8tWV6H+qJAQBAmSAYe8GWH0/r92JWiyXp9xyLOj3+P2X84ZluEw1rB2veje3Lbde31XuO6/8+SvLItWyzf/KGtoRiAABQZgjGXpB4+LRT4zwVivu0qqfXx3TxyLXcMeG/27Uu2XO10VHmIM0YElemLeUAAAAIxl5gNTxTFlGSns3q6pVbOys4wNcrz1eYWauTPBaKx14Ro36to8tt1RsAAFQvFeb/pZ966imZTCbdf//99mPZ2dmaNGmS6tSpo5o1a2r48OE6ceKEw+OOHTumQYMGKSQkRBEREZo6dary8jyz8uoptfzL9t8fwf4+Wji6o/47vmu5huInVu/X4k1HPXKtwW2jNeO6NopvWodQDAAAvKJCrBhv375dr7zyitq1a+dw/IEHHtAnn3yiFStWyGw2a/LkyRo2bJg2bdokSbJYLBo0aJCioqK0efNmpaam6rbbbpO/v7+efPLJ8ngphfrmR+dKKVwVEuCjV0Z11hXN6pZ7eJy7NlmvbzrikWsF+vnouZGXeeRaAAAAzir3FeOzZ89q1KhReu2111S7dm378czMTC1atEjPPvusevXqpU6dOmnx4sXavHmztmzZIkn63//+p+TkZC1dulQdOnTQtddeq8cff1wvvfSScnM9t0NcaWV6cHMOm7HxMUqefa16tKhX7qE4N8+qVzemeOx6829qX+6vCQAAVD/lHownTZqkQYMGqU+fPg7Hd+zYofPnzzscb9mypRo1aqTExERJUmJiotq2bavIyEj7mP79+ysrK0v79+8v8jlzcnKUlZXl8FWWokIDPXatEH+TXr7lMs0Y2sZj1yyt6Sv3ynB2c78S5G/cQY9iAADgfeVaSvHOO+9o586d2r59e4FzaWlpCggIUFhYmMPxyMhIpaWl2cdcGIpt523nijJ37lzNmjWrlLN3Xv820fr8+19LdY3LGobpof4t1K1Jxaq5XbP7uD7Y+UuprxNeI0BPDG3DbnYAAKDclFsw/umnn3Tfffdp3bp1CgoK8upzT58+XVOmTLF/n5WVpYYNG5bZ89U3B7v92CA/Hz17U/sKuYo6d22yXvm69CUUQ9pF6T8jOlaowA8AAKqfcgvGO3bs0MmTJ9WxY0f7MYvFoq+//lovvviiPvvsM+Xm5iojI8Nh1fjEiROKioqSJEVFRWnbtm0O17V1rbCNKUxgYKACAz1X3lAiN/Pe4HbRem7EZRUyMK7Z/UupQ7GPSZrQI1bTB8Z5aFYAAADuK7dg3Lt3b+3bt8/h2NixY9WyZUtNmzZNDRs2lL+/v9avX6/hw4dLkg4cOKBjx44pPj5ekhQfH685c+bo5MmTioiIkCStW7dOoaGhiourOGHrZFa2S+O7Na6t/97RrcLu8rZm93FNfme3W48NDfLT0A711bhODd0a37jCvkYAAFD9lFswrlWrltq0cbyBrEaNGqpTp479+Pjx4zVlyhSFh4crNDRU9957r+Lj49WtWzdJUr9+/RQXF6dbb71V8+bNU1pamh555BFNmjTJuyvCJTh11rkOGX1a1dPLozpX6LBY2vKJeTe2Ywc7AABQIVWIPsZFmT9/vnx8fDR8+HDl5OSof//+evnll+3nfX19tWbNGk2cOFHx8fGqUaOGxowZo9mzZ5fjrAv67XfngnHziNAKHYrX7k0tVSh+oE8zQjEAAKiwTIbhqUZblVdWVpbMZrMyMzMVGhrq8evf/84ufbT7eInjru9QX/8ZUbE2trBYDW1LSVda5h96dFWSzuZY3LpOtDlI30zrVSHrpQEAQMVX1nlNquArxlVFdJhzXTecHectCUmpmrU6WamZrtVIX8wkacaQOEIxAACo0AjGXhAe4ly9s7PjvGHt3lTds3xnqa8TbQ7SjCFxlFAAAIAKj2DsBXVrBnh0XFlbu/e4Jr+9q1TX6NWirib0vFRdYsNZKQYAAJUCwdgLopzc4MPZcWUpISlV9ywvXSgef2WsHh1ccdrlAQAAOINg7AVdYsMVbQ4qtlY32hykLrHhXpzVX+w32GVl6/E1+0t1rQk9YvV/gwjFAACg8iEYe4Gvj0nXtY8uttXZde2jy6XkwFM32NUM9NW84e0q5NbVAAAAziAYe4HFaujdb38udsy73/6sfwxo5dVwnJCUqolLd6q0/fr+b2ArjbsyllpiAABQqVXc3SSqkC2HTyvj3Plix2ScO68th097aUb5YX3W6uRSh2JJiggNJBQDAIBKj2DsBYk/nvLoOE/YlpJe6vIJm4haFav/MgAAgDsIxl7h7Gqq91ZdT57xTCguz5sGAQAAPIlg7AVdnQyOzo5zl8VqKPHwaa3a/YtOncnxyDXZ0Q4AAFQV3HznBT4m54Kjs+NcZbEaenHDQS3edEQZf/xV6+xjkqxuFhmbJL10S0d2tAMAAFUGwdgLTv3u3Oqss+NckZCUqodX7iv05j93Q7EkvXTLZRrYjlAMAACqDoKxF9StEejRcc5auzdV9yzf6dFrhoX466lhbVkpBgAAVQ7B2Bu8fO+dxWro+fU/6Pn1hzxzwT/dc3VTPdivBTXFAACgSiIYe8Gps06WUjg5rjjFlU6UVo9m9QjFAACgyiIYe4E3SilsN9jN//yg29coiklSFG3ZAABAFUcw9oYyLqVYuzdVj6xKUvrvue5dwAm0ZQMAAFUdwdgLyqqUwmI1dO/ynVqblObOtJwSXsNfT97AzXYAAKDqIxh7Qd2aTpZSODlOyl8lvu/dXTpvcb/nmklScY+uUyNAidN7K8CPfWAAAEDVR+LxAquTDYOdHTd3bbLuWb7T7VBsknR/72Z66ZaOMqlgBYft2Jwb2hCKAQBAtUHq8YKtKekeG7d273G98nVKqebz0i2X6f6+zTWwXbQWjO6oKHOQw/koc5AWjGZXOwAAUL1QSuEVzq7sFj/uj1yLHlyxx+1ZFLY5x4A20eobF6VtKek6eSZbEbXyu09wox0AAKhuCMZeEN+krl784rBT44oyd22yXv06xemIfSFzsJ/GdW+iyb0uLTTw+vqYFN+0jhtXBgAAqDoopfCCy53s/1vYOIvV0KRlO/SKm6H4/t7NtGV6H9UM9NWs1fu1aOOPys2zunElAACAqo0VYy/Y7mSN8faUdHVvlr9qnJtn1fQP9mjVnlTlOXlT3sXu6B6rP87nqfWMBF14iTlrv9OEHrGaPjDOresCAABURQRjL0j88ZTT47o3q6vH1+zXom+OlOo5+7SqJ19fFXqjntX46zjhGAAAIB+lFF5gOLngaxjS4Oe+KnUontAjVi+P6qzXNhbfveK1jSmUVQAAAPyJFWMvCAsJcGrcK18flrs51ddHur59fc0d3l6+PibN/DhJJVVgWA3prcQjGt+jiXtPCgAAUIUQjL2gbk3ngrG7odgkKWnmAAUH+CohKVUzP96vtCzntpc+mn7OvScFAACoYgjGXhBlDi7T69/ZM9Yeiu9eutOlx8aEh5TRrAAAACoXaoy9oEtsuKIv2l3OU2zdJSxWQw+v3OfSY31M0q3xjctkXgAAAJUNwdgLfH1MmjEkTp7eS+7FEZfp/wbld5XYcvi0Ms6dd+nxE3rEKsCPjwAAAIBEMPaaAW2idWfPWHlip+WQAB8tHN1RgzvUtx97a8sRpx/vY5Lu6kkfYwAAgAtRY+wlCUmpbm/pfKFBbaP1/MjLHLZ2nvPJfiXsP+HU46+8tI7euL0LK8UAAAAXIRh7gcVqaNbq5FKH4hdHXOawSixJcz5J1msbjzh9jYlXX0ooBgAAKATB2Au2paQrNTO7VNd4+ZbLNLCdYyheuze1xE08LlQ7xF/dmtQp1TwAAACqKoKxF/zyW+l6BS8c3VED2kQ7HLNYDT2yKsml68wd1tahBAMAAAB/IRh7QcL+VLceVzfET1sf6VdomH1h/Q9K/z3X6Ws90Kd5gXANAACAvxCMvSDNjTKKq1vU1ZKxXQs9N3dtsl752vkSivAQf03udanLcwAAAKhOCMZeYA52bktomwk9Yu39iS+2dm+qS6FYkp64nhIKAACAkhCMvaDHpXW16fDpEsdd07yuXrnt8iK7Rlishqa8t9ul557Qo7EGtqOEAgAAoCT07fKCSCe3g76uQ4NiW6nd984uZedZnX7ea1rU0/8Nau30eAAAgOqMYOwFUebgUo/LzbPqk72u3cR3Z8+mLo0HAACozgjGXtAlNlyBJWyqEejnoy6x4UWef3PzEZc2CIk2BxV7PQAAADgiGHtBbp5VOSWUQOTkWZVbzJg1e39x6TlnDInjhjsAAAAXEIy94Mm1yaUaZ7EaOnDirFPX8PUpfEMQAAAAFI+uFF5w5LRzO98VNW5bSrqyzzt3090bt12uq1pGOD03AAAA5GPF2Asa1wkp1biTZ5zbICTQz0dXNq/n9LwAAADwF4KxF/xzYOGbdTg7LqKWc+3e7rn6UuqKAQAA3EQw9oLgAF/1jSu+vKFvXISCA3zt31ushhIPn9aq3b/IahiKCg1UcZE3jG2fAQAASoUaYy9pUreG0+cTklI1a3WyUjP/KqEIC/GXIckkFdq27alhbPsMAABQGqwYe0FunlWvbUwpdsxrG1OUm2dVQlKqJi7d6RCKJSnz3HlJkjnE3+F4tDmILhQAAAAewIqxF7yVeETWEnbnsBrSm5tT9MKGQ4WuCNtWi4P9ffXS+I469XuOImrlb+LBSjEAAEDpEYy94MdTvzs1bl7CAZ0vJkEbklIzs+XjY9LQDg08NDsAAABIlFJ4xcks59qtFReKHa7nZPs2AAAAOI9g7AURoYGevZ6T7dsAAADgPIKxFzQKd26DD2eEhfirS2y4x64HAACAfARjLzCcq5BwytgrYrnZDgAAoAyUazBesGCB2rVrp9DQUIWGhio+Pl6ffvqp/Xx2drYmTZqkOnXqqGbNmho+fLhOnDjhcI1jx45p0KBBCgkJUUREhKZOnaq8vDxvv5Ri7Tia4ZHrmCQ28QAAACgj5RqML7nkEj311FPasWOHvv32W/Xq1UtDhw7V/v37JUkPPPCAVq9erRUrVuirr77S8ePHNWzYMPvjLRaLBg0apNzcXG3evFlvvvmmlixZoscee6y8XlKh/jjvmaA+pH00q8UAAABlxGQYnvyP/tILDw/XM888oxtvvFH16tXT8uXLdeONN0qSvv/+e7Vq1UqJiYnq1q2bPv30Uw0ePFjHjx9XZGSkJGnhwoWaNm2afv31VwUEBDj1nFlZWTKbzcrMzFRoaKjHX9PTa7/Tgq9/LPV1fnjiWgX4Uf0CAACqn7LOa1IFqjG2WCx655139Pvvvys+Pl47duzQ+fPn1adPH/uYli1bqlGjRkpMTJQkJSYmqm3btvZQLEn9+/dXVlaWfdW5MDk5OcrKynL4KktXNqtX6mtM6BFLKAYAAChD5Z609u3bp5o1ayowMFB33323PvzwQ8XFxSktLU0BAQEKCwtzGB8ZGam0tDRJUlpamkMotp23nSvK3LlzZTab7V8NGzb07Iu6SLemdRR20VbOrujVsp7+b1CcB2cEAACAi5V7MG7RooV2796trVu3auLEiRozZoySk5PL9DmnT5+uzMxM+9dPP/1Ups/n62PSU8PauvXYOiG+euP2Lh6eEQAAAC5W7ltCBwQE6NJL8zstdOrUSdu3b9dzzz2nm2++Wbm5ucrIyHBYNT5x4oSioqIkSVFRUdq2bZvD9WxdK2xjChMYGKjAQM9uulGSAW2i1SKyhg6ccG57aJvnR3YuoxkBAADgQuW+Ynwxq9WqnJwcderUSf7+/lq/fr393IEDB3Ts2DHFx8dLkuLj47Vv3z6dPHnSPmbdunUKDQ1VXFzFKj2wWA2XQ3Ggr0ndmtYpoxkBAADgQuW6Yjx9+nRde+21atSokc6cOaPly5fryy+/1GeffSaz2azx48drypQpCg8PV2hoqO69917Fx8erW7dukqR+/fopLi5Ot956q+bNm6e0tDQ98sgjmjRpktdXhEvy4oaDLj/mXzd1oD0bAACAl5RrMD558qRuu+02paamymw2q127dvrss8/Ut29fSdL8+fPl4+Oj4cOHKycnR/3799fLL79sf7yvr6/WrFmjiRMnKj4+XjVq1NCYMWM0e/bs8npJhbJYDb3iYru23i3rakj7+mU0IwAAAFyswvUxLg9l3Rdv08FTGrVoq9PjY8KD9dU/enl8HgAAAJVVtepjXJVt/vGU02NNkjY8dE3ZTQYAAACFIhh7wS+//eH02AWjO1JXDAAAUA4Ixt7gZLHK5Y1ra0Cb6LKdCwAAAApFMPaC+rWDnRp3eePwMp4JAAAAikIw9oLul9b16DgAAAB4HsHYCy5vHK6SqoZNYsUYAACgPBGMvWB7SnqJZcbGn+MAAABQPgjGXuBsuzZX2roBAADAswjGXnDcyXZtzo4DAACA5xGMvSCiVqBHxwEAAMDzCMZecCYnz6PjAAAA4HkEYy/49UyOR8cBAADA8wjGXnAu17mVYGfHAQAAwPMIxl4QFuTn0XEAAADwPIKxFxw+9btHxwEAAMDzCMZe8FO6c23YnB0HAAAAzyMYe8G581aPjgMAAIDnEYy9oKTtoF0dBwAAAM8jGAMAAAAiGAMAAACSCMYAAACAJIKxV5g8PA4AAACeRzD2ggBfz44DAACA5xGMvSAsJMCj4wAAAOB5BGMveOK6Nh4dBwAAAM8jGHtBr9ZRJdYPm/4cBwAAgPJBMPYCXx+TFozuWOyYBaM7yteH2+8AAADKC8HYSwa0idbC0R1Vr4afw/F6Nfy1cHRHDWgTXU4zAwAAgCT5lTwEnjKgTbT6xkVpW0q6Tp7JVkStIHWJDWelGAAAoAIgGHuZr49J8U3rlPc0AAAAcBFKKQAAAAARjAEAAABJBGMAAABAEsEYAAAAkEQwBgAAACQRjAEAAABJBGMAAABAEsEYAAAAkEQwBgAAACQRjAEAAABJBGMAAABAEsEYAAAAkEQwBgAAACRJfuU9gYrAMAxJUlZWVjnPBAAAAIWx5TRbbisLBGNJZ86ckSQ1bNiwnGcCAACA4pw5c0Zms7lMrm0yyjJ2VxJWq1XHjx9XrVq1ZDKZyvz5srKy1LBhQ/30008KDQ0t8+erjHiPnMP75Bzep5LxHjmH98k5vE8l4z1yzoXvU61atXTmzBnVr19fPj5lUw3MirEkHx8fXXLJJV5/3tDQUP4wlID3yDm8T87hfSoZ75FzeJ+cw/tUMt4j59jep7JaKbbh5jsAAABABGMAAABAEsG4XAQGBmrGjBkKDAws76lUWLxHzuF9cg7vU8l4j5zD++Qc3qeS8R45x9vvEzffAQAAAGLFGAAAAJBEMAYAAAAkEYwBAAAASQRjAAAAQBLB2OteeuklNW7cWEFBQeratau2bdtW3lPymrlz5+ryyy9XrVq1FBERoeuvv14HDhxwGHP11VfLZDI5fN19990OY44dO6ZBgwYpJCREERERmjp1qvLy8rz5UsrUzJkzC7wHLVu2tJ/Pzs7WpEmTVKdOHdWsWVPDhw/XiRMnHK5R1d8jSWrcuHGB98lkMmnSpEmSqudn6euvv9aQIUNUv359mUwmffTRRw7nDcPQY489pujoaAUHB6tPnz46ePCgw5j09HSNGjVKoaGhCgsL0/jx43X27FmHMXv37lWPHj0UFBSkhg0bat68eWX90jyquPfp/PnzmjZtmtq2basaNWqofv36uu2223T8+HGHaxT2+XvqqaccxlTl90mSbr/99gLvwYABAxzGVPXPU0nvUWF/R5lMJj3zzDP2MVX9s+TMz35P/Vz78ssv1bFjRwUGBurSSy/VkiVLXJ+wAa955513jICAAOONN94w9u/fb0yYMMEICwszTpw4Ud5T84r+/fsbixcvNpKSkozdu3cbAwcONBo1amScPXvWPuaqq64yJkyYYKSmptq/MjMz7efz8vKMNm3aGH369DF27dplrF271qhbt64xffr08nhJZWLGjBlG69atHd6DX3/91X7+7rvvNho2bGisX7/e+Pbbb41u3boZV1xxhf18dXiPDMMwTp486fAerVu3zpBkfPHFF4ZhVM/P0tq1a43/+7//M1auXGlIMj788EOH80899ZRhNpuNjz76yNizZ49x3XXXGbGxscYff/xhHzNgwACjffv2xpYtW4yNGzcal156qTFy5Ej7+czMTCMyMtIYNWqUkZSUZLz99ttGcHCw8corr3jrZZZace9TRkaG0adPH+Pdd981vv/+eyMxMdHo0qWL0alTJ4drxMTEGLNnz3b4fF34d1lVf58MwzDGjBljDBgwwOE9SE9PdxhT1T9PJb1HF743qampxhtvvGGYTCbj8OHD9jFV/bPkzM9+T/xc+/HHH42QkBBjypQpRnJysvHCCy8Yvr6+RkJCgkvzJRh7UZcuXYxJkybZv7dYLEb9+vWNuXPnluOsys/JkycNScZXX31lP3bVVVcZ9913X5GPWbt2reHj42OkpaXZjy1YsMAIDQ01cnJyynK6XjNjxgyjffv2hZ7LyMgw/P39jRUrVtiPfffdd4YkIzEx0TCM6vEeFea+++4zmjZtalitVsMw+Cxd/EPaarUaUVFRxjPPPGM/lpGRYQQGBhpvv/22YRiGkZycbEgytm/fbh/z6aefGiaTyfjll18MwzCMl19+2ahdu7bDezRt2jSjRYsWZfyKykZhYeZi27ZtMyQZR48etR+LiYkx5s+fX+RjqsP7NGbMGGPo0KFFPqa6fZ6c+SwNHTrU6NWrl8Ox6vZZuvhnv6d+rv3jH/8wWrdu7fBcN998s9G/f3+X5kcphZfk5uZqx44d6tOnj/2Yj4+P+vTpo8TExHKcWfnJzMyUJIWHhzscX7ZsmerWras2bdpo+vTpOnfunP1cYmKi2rZtq8jISPux/v37KysrS/v37/fOxL3g4MGDql+/vpo0aaJRo0bp2LFjkqQdO3bo/PnzDp+jli1bqlGjRvbPUXV5jy6Um5urpUuXaty4cTKZTPbjfJb+kpKSorS0NIfPjtlsVteuXR0+O2FhYercubN9TJ8+feTj46OtW7fax/Ts2VMBAQH2Mf3799eBAwf022+/eenVeFdmZqZMJpPCwsIcjj/11FOqU6eOLrvsMj3zzDMO/61bXd6nL7/8UhEREWrRooUmTpyo06dP28/xeXJ04sQJffLJJxo/fnyBc9Xps3Txz35P/VxLTEx0uIZtjKsZy8/1lwR3nDp1ShaLxeE3VZIiIyP1/fffl9Osyo/VatX999+v7t27q02bNvbjt9xyi2JiYlS/fn3t3btX06ZN04EDB7Ry5UpJUlpaWqHvoe1cVdC1a1ctWbJELVq0UGpqqmbNmqUePXooKSlJaWlpCggIKPADOjIy0v76q8N7dLGPPvpIGRkZuv322+3H+Cw5sr2mwl7zhZ+diIgIh/N+fn4KDw93GBMbG1vgGrZztWvXLpP5l5fs7GxNmzZNI0eOVGhoqP343//+d3Xs2FHh4eHavHmzpk+frtTUVD377LOSqsf7NGDAAA0bNkyxsbE6fPiw/vnPf+raa69VYmKifH19+Txd5M0331StWrU0bNgwh+PV6bNU2M9+T/1cK2pMVlaW/vjjDwUHBzs1R4IxysWkSZOUlJSkb775xuH4nXfeaf9127ZtFR0drd69e+vw4cNq2rSpt6dZLq699lr7r9u1a6euXbsqJiZG7733ntN/sKubRYsW6dprr1X9+vXtx/gsobTOnz+vm266SYZhaMGCBQ7npkyZYv91u3btFBAQoLvuuktz586tNlv8jhgxwv7rtm3bql27dmratKm+/PJL9e7duxxnVjG98cYbGjVqlIKCghyOV6fPUlE/+ysSSim8pG7duvL19S1wl+WJEycUFRVVTrMqH5MnT9aaNWv0xRdf6JJLLil2bNeuXSVJhw4dkiRFRUUV+h7azlVFYWFhat68uQ4dOqSoqCjl5uYqIyPDYcyFn6Pq9h4dPXpUn3/+ue64445ix1X3z5LtNRX3d1BUVJROnjzpcD4vL0/p6enV7vNlC8VHjx7VunXrHFaLC9O1a1fl5eXpyJEjkqrP+3ShJk2aqG7dug5/xvg85du4caMOHDhQ4t9TUtX9LBX1s99TP9eKGhMaGurSohLB2EsCAgLUqVMnrV+/3n7MarVq/fr1io+PL8eZeY9hGJo8ebI+/PBDbdiwocB/DRVm9+7dkqTo6GhJUnx8vPbt2+fwl63th1ZcXFyZzLu8nT17VocPH1Z0dLQ6deokf39/h8/RgQMHdOzYMfvnqLq9R4sXL1ZERIQGDRpU7Ljq/lmKjY1VVFSUw2cnKytLW7dudfjsZGRkaMeOHfYxGzZskNVqtf/DIj4+Xl9//bXOnz9vH7Nu3Tq1aNGiUv2XbnFsofjgwYP6/PPPVadOnRIfs3v3bvn4+NhLB6rD+3Sxn3/+WadPn3b4M8bnKd+iRYvUqVMntW/fvsSxVe2zVNLPfk/9XIuPj3e4hm2MyxnL9fsJ4a533nnHCAwMNJYsWWIkJycbd955pxEWFuZwl2VVNnHiRMNsNhtffvmlQ1uac+fOGYZhGIcOHTJmz55tfPvtt0ZKSoqxatUqo0mTJkbPnj3t17C1bOnXr5+xe/duIyEhwahXr16lbrF1sQcffND48ssvjZSUFGPTpk1Gnz59jLp16xonT540DCO/rU2jRo2MDRs2GN9++60RHx9vxMfH2x9fHd4jG4vFYjRq1MiYNm2aw/Hq+lk6c+aMsWvXLmPXrl2GJOPZZ581du3aZe+m8NRTTxlhYWHGqlWrjL179xpDhw4ttF3bZZddZmzdutX45ptvjGbNmjm018rIyDAiIyONW2+91UhKSjLeeecdIyQkpNK0jjKM4t+n3Nxc47rrrjMuueQSY/fu3Q5/V9nuft+8ebMxf/58Y/fu3cbhw4eNpUuXGvXq1TNuu+02+3NU9ffpzJkzxkMPPWQkJiYaKSkpxueff2507NjRaNasmZGdnW2/RlX/PJX0Z84w8tuthYSEGAsWLCjw+OrwWSrpZ79heObnmq1d29SpU43vvvvOeOmll2jXVhm88MILRqNGjYyAgACjS5cuxpYtW8p7Sl4jqdCvxYsXG4ZhGMeOHTN69uxphIeHG4GBgcall15qTJ061aH3rGEYxpEjR4xrr73WCA4ONurWrWs8+OCDxvnz58vhFZWNm2++2YiOjjYCAgKMBg0aGDfffLNx6NAh+/k//vjDuOeee4zatWsbISEhxg033GCkpqY6XKOqv0c2n332mSHJOHDggMPx6vpZ+uKLLwr9MzZmzBjDMPJbtj366KNGZGSkERgYaPTu3bvAe3f69Glj5MiRRs2aNY3Q0FBj7NixxpkzZxzG7Nmzx7jyyiuNwMBAo0GDBsZTTz3lrZfoEcW9TykpKUX+XWXrkb1jxw6ja9euhtlsNoKCgoxWrVoZTz75pEMgNIyq/T6dO3fO6Nevn1GvXj3D39/fiImJMSZMmFBgoaeqf55K+jNnGIbxyiuvGMHBwUZGRkaBx1eHz1JJP/sNw3M/17744gujQ4cORkBAgNGkSROH53CW6c9JAwAAANUaNcYAAACACMYAAACAJIIxAAAAIIlgDAAAAEgiGAMAAACSCMYAAACAJIIxAAAAIIlgDAAAAEgiGANAtWEymfTRRx+V9zQAoMIiGAOAhyUmJsrX11eDBg1y+bGNGzfWf/7zH89PyklpaWm699571aRJEwUGBqphw4YaMmSI1q9fX+wchwwZogEDBhR6zY0bN8pkMmnv3r1lOXUAKDWCMQB42KJFi3Tvvffq66+/1vHjx8t7Ok47cuSIOnXqpA0bNuiZZ57Rvn37lJCQoGuuuUaTJk0q9rHjx4/XunXr9PPPPxc4t3jxYnXu3Fnt2rUrq6kDgEcQjAHAg86ePat3331XEydO1KBBg7RkyZICY1avXq3LL79cQUFBqlu3rm644QZJ0tVXX62jR4/qgQcekMlkkslkkiTNnDlTHTp0cLjGf/7zHzVu3Nj+/fbt29W3b1/VrVtXZrNZV111lXbu3OnS3O+55x6ZTCZt27ZNw4cPV/PmzdW6dWtNmTJFW7ZsKfaxgwcPVr169Qq83rNnz2rFihUaP368S3MBgPJAMAYAD3rvvffUsmVLtWjRQqNHj9Ybb7whwzDs5z/55BPdcMMNGjhwoHbt2qX169erS5cukqSVK1fqkksu0ezZs5WamqrU1FSnn/fMmTMaM2aMvvnmG23ZskXNmjXTwIEDdebMGacen56eroSEBE2aNEk1atQocD4sLKzYx/v5+em2227TkiVLHF7vihUrZLFYNHLkSKdfCwCUF7/yngAAVCWLFi3S6NGjJUkDBgxQZmamvvrqK1199dWSpDlz5mjEiBGaNWuW/THt27eXJIWHh8vX11e1atVSVFSUS8/bq1cvh+9fffVVhYWF6auvvtLgwYNLfPyhQ4dkGIZatmzp0vNeaNy4cXrmmWccXu/ixYs1fPhwmc1mt68LAN7CijEAeMiBAwe0bds2++qon5+fbr75Zi1atMg+Zvfu3erdu7fHn/vEiROaMGGCmjVrJrPZrNDQUJ09e1bHjh1z6vEXrvK6q2XLlrriiiv0xhtvSMoP2xs3bqSMAkClwYoxAHjIokWLlJeXp/r169uPGYahwMBAvfjiizKbzQoODnb5uj4+PgWC6/nz5x2+HzNmjE6fPq3nnntOMTExCgwMVHx8vHJzc516jmbNmslkMun77793eX4XGj9+vO6991699NJLWrx4sZo2baqrrrqqVNcEAG9hxRgAPCAvL0///e9/9e9//1u7d++2f+3Zs0f169fX22+/LUlq166dQ+uziwUEBMhisTgcq1evntLS0hzC8e7dux3GbNq0SX//+981cOBAtW7dWoGBgTp16pTT8w8PD1f//v310ksv6ffffy9wPiMjw6nr3HTTTfLx8dHy5cv13//+V+PGjbPfRAgAFR0rxgDgAWvWrNFvv/2m8ePHF6inHT58uBYtWqS7775bM2bMUO/evdW0aVONGDFCeXl5Wrt2raZNmyYpv0fw119/rREjRigwMFB169bV1VdfrV9//VXz5s3TjTfeqISEBH366acKDQ21P0ezZs301ltvqXPnzsrKytLUqVNdXp1+6aWX1L17d3Xp0kWzZ89Wu3btlJeXp3Xr1mnBggX67rvv7GN/+eWXAuE8JiZGtWvX1s0336zp06crKytLt99+u2tvJACUJwMAUGqDBw82Bg4cWOi5rVu3GpKMPXv2GIZhGB988IHRoUMHIyAgwKhbt64xbNgw+9jExESjXbt2RmBgoHHhX9ELFiwwGjZsaNSoUcO47bbbjDlz5hgxMTH28zt37jQ6d+5sBAUFGc2aNTNWrFhhxMTEGPPnz7ePkWR8+OGHxb6O48ePG5MmTTJiYmKMgIAAo0GDBsZ1111nfPHFF/YxMTExhqQCX2+99ZZhGIaxefNmQ1KR7wcAVFQmw/DAHRcAAABAJUeNMQAAACCCMQAAACCJYAwAAABIIhgDAAAAkgjGAAAAgCSCMQAAACCJYAwAAABIIhgDAAAAkgjGAAAAgCSCMQAAACCJYAwAAABIkv4fUNpqhzBCRu8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.figure(figsize=(8, 6))\n",
    "plt.scatter(y_test, y_pred)\n",
    "plt.xlabel('Actual CLV')\n",
    "plt.ylabel('Predicted CLV')\n",
    "plt.title('Actual vs. Predicted CLV')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jzQrwyYR9uN7"
   },
   "source": [
    "<h2>6. Scaling up: developing a model that overfits</h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ZGlKltml9yQB",
    "outputId": "200c235e-13ad-433a-9bd5-17b294ad4ffc"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/200\n",
      "45/45 [==============================] - 7s 71ms/step - loss: 87241.1406 - mean_absolute_error: 182.2564 - val_loss: 1855.2557 - val_mean_absolute_error: 29.7510\n",
      "Epoch 2/200\n",
      "45/45 [==============================] - 3s 65ms/step - loss: 6725.9707 - mean_absolute_error: 57.4166 - val_loss: 964.6232 - val_mean_absolute_error: 20.6391\n",
      "Epoch 3/200\n",
      "45/45 [==============================] - 3s 74ms/step - loss: 3698.3113 - mean_absolute_error: 41.8332 - val_loss: 741.2234 - val_mean_absolute_error: 19.6441\n",
      "Epoch 4/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 2233.6169 - mean_absolute_error: 32.6165 - val_loss: 500.2681 - val_mean_absolute_error: 16.8761\n",
      "Epoch 5/200\n",
      "45/45 [==============================] - 4s 81ms/step - loss: 1410.6387 - mean_absolute_error: 25.6009 - val_loss: 242.9864 - val_mean_absolute_error: 10.1537\n",
      "Epoch 6/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 1033.5745 - mean_absolute_error: 21.4800 - val_loss: 330.4153 - val_mean_absolute_error: 13.9001\n",
      "Epoch 7/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 892.9293 - mean_absolute_error: 20.4044 - val_loss: 247.3096 - val_mean_absolute_error: 10.2307\n",
      "Epoch 8/200\n",
      "45/45 [==============================] - 2s 52ms/step - loss: 721.9688 - mean_absolute_error: 16.8559 - val_loss: 198.0465 - val_mean_absolute_error: 9.5584\n",
      "Epoch 9/200\n",
      "45/45 [==============================] - 2s 51ms/step - loss: 555.8040 - mean_absolute_error: 15.0489 - val_loss: 139.4440 - val_mean_absolute_error: 7.9589\n",
      "Epoch 10/200\n",
      "45/45 [==============================] - 3s 59ms/step - loss: 332.2649 - mean_absolute_error: 11.3507 - val_loss: 117.9194 - val_mean_absolute_error: 7.3301\n",
      "Epoch 11/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 280.7074 - mean_absolute_error: 10.3052 - val_loss: 112.3908 - val_mean_absolute_error: 6.8798\n",
      "Epoch 12/200\n",
      "45/45 [==============================] - 3s 69ms/step - loss: 194.6300 - mean_absolute_error: 8.9255 - val_loss: 58.9626 - val_mean_absolute_error: 4.6275\n",
      "Epoch 13/200\n",
      "45/45 [==============================] - 2s 53ms/step - loss: 110.2101 - mean_absolute_error: 6.1728 - val_loss: 45.9236 - val_mean_absolute_error: 4.0382\n",
      "Epoch 14/200\n",
      "45/45 [==============================] - 3s 67ms/step - loss: 109.0466 - mean_absolute_error: 6.4272 - val_loss: 40.0352 - val_mean_absolute_error: 3.7968\n",
      "Epoch 15/200\n",
      "45/45 [==============================] - 3s 75ms/step - loss: 72.1114 - mean_absolute_error: 5.0850 - val_loss: 44.3621 - val_mean_absolute_error: 4.7498\n",
      "Epoch 16/200\n",
      "45/45 [==============================] - 3s 77ms/step - loss: 71.3919 - mean_absolute_error: 5.3394 - val_loss: 32.3621 - val_mean_absolute_error: 3.7358\n",
      "Epoch 17/200\n",
      "45/45 [==============================] - 3s 67ms/step - loss: 58.8237 - mean_absolute_error: 4.6830 - val_loss: 73.6778 - val_mean_absolute_error: 6.6006\n",
      "Epoch 18/200\n",
      "45/45 [==============================] - 2s 53ms/step - loss: 46.8374 - mean_absolute_error: 4.2898 - val_loss: 34.1491 - val_mean_absolute_error: 4.3761\n",
      "Epoch 19/200\n",
      "45/45 [==============================] - 3s 67ms/step - loss: 29.0161 - mean_absolute_error: 3.2828 - val_loss: 16.5235 - val_mean_absolute_error: 2.6760\n",
      "Epoch 20/200\n",
      "45/45 [==============================] - 3s 76ms/step - loss: 22.3717 - mean_absolute_error: 2.9506 - val_loss: 20.4074 - val_mean_absolute_error: 3.0962\n",
      "Epoch 21/200\n",
      "45/45 [==============================] - 2s 50ms/step - loss: 24.9142 - mean_absolute_error: 3.1895 - val_loss: 16.0976 - val_mean_absolute_error: 2.4124\n",
      "Epoch 22/200\n",
      "45/45 [==============================] - 2s 51ms/step - loss: 17.5324 - mean_absolute_error: 2.6183 - val_loss: 10.0758 - val_mean_absolute_error: 1.7552\n",
      "Epoch 23/200\n",
      "45/45 [==============================] - 3s 60ms/step - loss: 14.6770 - mean_absolute_error: 2.4639 - val_loss: 10.8417 - val_mean_absolute_error: 1.7990\n",
      "Epoch 24/200\n",
      "45/45 [==============================] - 3s 58ms/step - loss: 17.8724 - mean_absolute_error: 2.7323 - val_loss: 10.5275 - val_mean_absolute_error: 2.1885\n",
      "Epoch 25/200\n",
      "45/45 [==============================] - 3s 75ms/step - loss: 11.6701 - mean_absolute_error: 2.2512 - val_loss: 15.0363 - val_mean_absolute_error: 2.6262\n",
      "Epoch 26/200\n",
      "45/45 [==============================] - 3s 60ms/step - loss: 9.2794 - mean_absolute_error: 2.0103 - val_loss: 15.8266 - val_mean_absolute_error: 3.0749\n",
      "Epoch 27/200\n",
      "45/45 [==============================] - 2s 51ms/step - loss: 10.3076 - mean_absolute_error: 2.0540 - val_loss: 21.4191 - val_mean_absolute_error: 3.6121\n",
      "Epoch 28/200\n",
      "45/45 [==============================] - 3s 67ms/step - loss: 12.6388 - mean_absolute_error: 2.4463 - val_loss: 11.6240 - val_mean_absolute_error: 2.3794\n",
      "Epoch 29/200\n",
      "45/45 [==============================] - 3s 62ms/step - loss: 5.4848 - mean_absolute_error: 1.4446 - val_loss: 7.4525 - val_mean_absolute_error: 1.5545\n",
      "Epoch 30/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 7.2381 - mean_absolute_error: 1.7040 - val_loss: 6.9166 - val_mean_absolute_error: 1.1614\n",
      "Epoch 31/200\n",
      "45/45 [==============================] - 3s 62ms/step - loss: 7.5064 - mean_absolute_error: 1.8586 - val_loss: 6.5513 - val_mean_absolute_error: 1.3413\n",
      "Epoch 32/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 8.3970 - mean_absolute_error: 1.8762 - val_loss: 11.7933 - val_mean_absolute_error: 2.5544\n",
      "Epoch 33/200\n",
      "45/45 [==============================] - 3s 58ms/step - loss: 10.3521 - mean_absolute_error: 2.1786 - val_loss: 8.7066 - val_mean_absolute_error: 1.9164\n",
      "Epoch 34/200\n",
      "45/45 [==============================] - 5s 103ms/step - loss: 11.1893 - mean_absolute_error: 2.2065 - val_loss: 7.6828 - val_mean_absolute_error: 1.6787\n",
      "Epoch 35/200\n",
      "45/45 [==============================] - 3s 60ms/step - loss: 7.6981 - mean_absolute_error: 1.7445 - val_loss: 10.0488 - val_mean_absolute_error: 1.9436\n",
      "Epoch 36/200\n",
      "45/45 [==============================] - 4s 88ms/step - loss: 10.2473 - mean_absolute_error: 2.1912 - val_loss: 8.2147 - val_mean_absolute_error: 1.5806\n",
      "Epoch 37/200\n",
      "45/45 [==============================] - 3s 65ms/step - loss: 7.0927 - mean_absolute_error: 1.8097 - val_loss: 6.3383 - val_mean_absolute_error: 1.2960\n",
      "Epoch 38/200\n",
      "45/45 [==============================] - 3s 67ms/step - loss: 3.5662 - mean_absolute_error: 1.2694 - val_loss: 5.6281 - val_mean_absolute_error: 0.9451\n",
      "Epoch 39/200\n",
      "45/45 [==============================] - 3s 56ms/step - loss: 2.4142 - mean_absolute_error: 1.0158 - val_loss: 5.4110 - val_mean_absolute_error: 0.9362\n",
      "Epoch 40/200\n",
      "45/45 [==============================] - 2s 51ms/step - loss: 2.5684 - mean_absolute_error: 1.0295 - val_loss: 7.8492 - val_mean_absolute_error: 1.6794\n",
      "Epoch 41/200\n",
      "45/45 [==============================] - 2s 52ms/step - loss: 3.3796 - mean_absolute_error: 1.2486 - val_loss: 6.9397 - val_mean_absolute_error: 1.4394\n",
      "Epoch 42/200\n",
      "45/45 [==============================] - 3s 59ms/step - loss: 4.0690 - mean_absolute_error: 1.3095 - val_loss: 6.3912 - val_mean_absolute_error: 1.3355\n",
      "Epoch 43/200\n",
      "45/45 [==============================] - 4s 78ms/step - loss: 4.4204 - mean_absolute_error: 1.5372 - val_loss: 7.4061 - val_mean_absolute_error: 1.7506\n",
      "Epoch 44/200\n",
      "45/45 [==============================] - 3s 59ms/step - loss: 2.5089 - mean_absolute_error: 1.1124 - val_loss: 6.3707 - val_mean_absolute_error: 1.5542\n",
      "Epoch 45/200\n",
      "45/45 [==============================] - 2s 53ms/step - loss: 2.6690 - mean_absolute_error: 1.1286 - val_loss: 6.5260 - val_mean_absolute_error: 1.3300\n",
      "Epoch 46/200\n",
      "45/45 [==============================] - 3s 63ms/step - loss: 1.9600 - mean_absolute_error: 0.9624 - val_loss: 5.8049 - val_mean_absolute_error: 1.3244\n",
      "Epoch 47/200\n",
      "45/45 [==============================] - 3s 69ms/step - loss: 3.0421 - mean_absolute_error: 1.1266 - val_loss: 5.7412 - val_mean_absolute_error: 1.4527\n",
      "Epoch 48/200\n",
      "45/45 [==============================] - 3s 60ms/step - loss: 103.1451 - mean_absolute_error: 5.9310 - val_loss: 53.1362 - val_mean_absolute_error: 5.0147\n",
      "Epoch 49/200\n",
      "45/45 [==============================] - 2s 53ms/step - loss: 132.8705 - mean_absolute_error: 7.7631 - val_loss: 11.3476 - val_mean_absolute_error: 2.2520\n",
      "Epoch 50/200\n",
      "45/45 [==============================] - 2s 52ms/step - loss: 15.1604 - mean_absolute_error: 2.8121 - val_loss: 14.2328 - val_mean_absolute_error: 2.5142\n",
      "Epoch 51/200\n",
      "45/45 [==============================] - 3s 61ms/step - loss: 12.0728 - mean_absolute_error: 2.4181 - val_loss: 8.2459 - val_mean_absolute_error: 1.5704\n",
      "Epoch 52/200\n",
      "45/45 [==============================] - 3s 67ms/step - loss: 11.5544 - mean_absolute_error: 2.4259 - val_loss: 12.6752 - val_mean_absolute_error: 2.3280\n",
      "Epoch 53/200\n",
      "45/45 [==============================] - 3s 65ms/step - loss: 4.7262 - mean_absolute_error: 1.4572 - val_loss: 7.2637 - val_mean_absolute_error: 1.5641\n",
      "Epoch 54/200\n",
      "45/45 [==============================] - 3s 61ms/step - loss: 6.1133 - mean_absolute_error: 1.7511 - val_loss: 7.8743 - val_mean_absolute_error: 1.9087\n",
      "Epoch 55/200\n",
      "45/45 [==============================] - 3s 61ms/step - loss: 3.3979 - mean_absolute_error: 1.2752 - val_loss: 5.2386 - val_mean_absolute_error: 1.0221\n",
      "Epoch 56/200\n",
      "45/45 [==============================] - 2s 53ms/step - loss: 1.7167 - mean_absolute_error: 0.8797 - val_loss: 5.2532 - val_mean_absolute_error: 1.1330\n",
      "Epoch 57/200\n",
      "45/45 [==============================] - 4s 82ms/step - loss: 1.8789 - mean_absolute_error: 0.9344 - val_loss: 7.7594 - val_mean_absolute_error: 1.9602\n",
      "Epoch 58/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 4.0295 - mean_absolute_error: 1.3971 - val_loss: 7.2010 - val_mean_absolute_error: 1.7867\n",
      "Epoch 59/200\n",
      "45/45 [==============================] - 3s 66ms/step - loss: 7.0578 - mean_absolute_error: 1.8391 - val_loss: 5.7910 - val_mean_absolute_error: 1.3520\n",
      "Epoch 60/200\n",
      "45/45 [==============================] - 3s 57ms/step - loss: 4.0161 - mean_absolute_error: 1.4024 - val_loss: 6.1679 - val_mean_absolute_error: 1.5477\n",
      "Epoch 61/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 2.3327 - mean_absolute_error: 1.0759 - val_loss: 4.6813 - val_mean_absolute_error: 0.9953\n",
      "Epoch 62/200\n",
      "45/45 [==============================] - 3s 66ms/step - loss: 3.3243 - mean_absolute_error: 1.1984 - val_loss: 4.6856 - val_mean_absolute_error: 0.8376\n",
      "Epoch 63/200\n",
      "45/45 [==============================] - 3s 64ms/step - loss: 4.0459 - mean_absolute_error: 1.3606 - val_loss: 7.0191 - val_mean_absolute_error: 1.8484\n",
      "Epoch 64/200\n",
      "45/45 [==============================] - 3s 60ms/step - loss: 13.2647 - mean_absolute_error: 2.4628 - val_loss: 6.8458 - val_mean_absolute_error: 1.6997\n",
      "Epoch 65/200\n",
      "45/45 [==============================] - 2s 52ms/step - loss: 17.9595 - mean_absolute_error: 2.8575 - val_loss: 7.4114 - val_mean_absolute_error: 1.8183\n",
      "Epoch 66/200\n",
      "45/45 [==============================] - 4s 82ms/step - loss: 25.3139 - mean_absolute_error: 3.2932 - val_loss: 7.5392 - val_mean_absolute_error: 1.6389\n",
      "Epoch 67/200\n",
      "45/45 [==============================] - 3s 71ms/step - loss: 6.5329 - mean_absolute_error: 1.7685 - val_loss: 5.0338 - val_mean_absolute_error: 1.1032\n",
      "Epoch 68/200\n",
      "45/45 [==============================] - 2s 54ms/step - loss: 109.6590 - mean_absolute_error: 6.9283 - val_loss: 74.9487 - val_mean_absolute_error: 6.7200\n",
      "Epoch 69/200\n",
      "45/45 [==============================] - 2s 53ms/step - loss: 65.3817 - mean_absolute_error: 5.5742 - val_loss: 14.4241 - val_mean_absolute_error: 2.6496\n",
      "Epoch 70/200\n",
      "45/45 [==============================] - 3s 58ms/step - loss: 8.3257 - mean_absolute_error: 1.9308 - val_loss: 6.2877 - val_mean_absolute_error: 1.5413\n",
      "Epoch 71/200\n",
      "45/45 [==============================] - 3s 74ms/step - loss: 10.8294 - mean_absolute_error: 2.1327 - val_loss: 4.8923 - val_mean_absolute_error: 1.1201\n",
      "Epoch 72/200\n",
      "45/45 [==============================] - 5s 104ms/step - loss: 4.6710 - mean_absolute_error: 1.4099 - val_loss: 4.3411 - val_mean_absolute_error: 0.9493\n",
      "Epoch 73/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 3.0649 - mean_absolute_error: 1.1733 - val_loss: 4.2255 - val_mean_absolute_error: 0.7965\n",
      "Epoch 74/200\n",
      "45/45 [==============================] - 3s 69ms/step - loss: 3.4633 - mean_absolute_error: 1.1341 - val_loss: 6.6289 - val_mean_absolute_error: 1.3327\n",
      "Epoch 75/200\n",
      "45/45 [==============================] - 3s 71ms/step - loss: 3.4342 - mean_absolute_error: 1.2320 - val_loss: 5.1079 - val_mean_absolute_error: 1.2413\n",
      "Epoch 76/200\n",
      "45/45 [==============================] - 3s 66ms/step - loss: 6.3204 - mean_absolute_error: 1.7011 - val_loss: 6.9843 - val_mean_absolute_error: 1.7520\n",
      "Epoch 77/200\n",
      "45/45 [==============================] - 4s 81ms/step - loss: 17.7025 - mean_absolute_error: 2.6370 - val_loss: 8.4618 - val_mean_absolute_error: 1.8106\n",
      "Epoch 78/200\n",
      "45/45 [==============================] - 4s 80ms/step - loss: 6.6988 - mean_absolute_error: 1.6773 - val_loss: 6.1210 - val_mean_absolute_error: 1.3721\n",
      "Epoch 79/200\n",
      "45/45 [==============================] - 3s 76ms/step - loss: 16.6383 - mean_absolute_error: 2.7325 - val_loss: 5.2243 - val_mean_absolute_error: 1.0809\n",
      "Epoch 80/200\n",
      "45/45 [==============================] - 3s 71ms/step - loss: 11.2248 - mean_absolute_error: 2.1589 - val_loss: 7.0860 - val_mean_absolute_error: 1.8870\n",
      "Epoch 81/200\n",
      "45/45 [==============================] - 3s 62ms/step - loss: 43.3928 - mean_absolute_error: 4.2271 - val_loss: 100.5794 - val_mean_absolute_error: 8.2832\n",
      "Epoch 82/200\n",
      "45/45 [==============================] - 4s 79ms/step - loss: 83.9403 - mean_absolute_error: 6.4423 - val_loss: 12.6122 - val_mean_absolute_error: 2.5685\n",
      "Epoch 83/200\n",
      "45/45 [==============================] - 3s 65ms/step - loss: 18.9957 - mean_absolute_error: 3.0208 - val_loss: 6.6682 - val_mean_absolute_error: 1.4620\n",
      "Epoch 84/200\n",
      "45/45 [==============================] - 3s 63ms/step - loss: 9.6205 - mean_absolute_error: 2.0561 - val_loss: 5.9963 - val_mean_absolute_error: 1.2691\n",
      "Epoch 85/200\n",
      "45/45 [==============================] - 2s 54ms/step - loss: 48.4690 - mean_absolute_error: 4.3983 - val_loss: 11.9859 - val_mean_absolute_error: 2.5631\n",
      "Epoch 86/200\n",
      "45/45 [==============================] - 3s 58ms/step - loss: 66.9818 - mean_absolute_error: 5.2818 - val_loss: 51.6477 - val_mean_absolute_error: 4.3479\n",
      "Epoch 87/200\n",
      "45/45 [==============================] - 4s 83ms/step - loss: 235.4227 - mean_absolute_error: 9.2824 - val_loss: 18.1418 - val_mean_absolute_error: 2.9553\n",
      "Epoch 88/200\n",
      "45/45 [==============================] - 3s 70ms/step - loss: 13.2052 - mean_absolute_error: 2.4102 - val_loss: 8.9439 - val_mean_absolute_error: 1.6139\n",
      "Epoch 89/200\n",
      "45/45 [==============================] - 3s 69ms/step - loss: 8.5351 - mean_absolute_error: 1.8481 - val_loss: 9.4460 - val_mean_absolute_error: 1.8383\n",
      "Epoch 90/200\n",
      "45/45 [==============================] - 3s 70ms/step - loss: 10.6047 - mean_absolute_error: 2.0076 - val_loss: 6.0010 - val_mean_absolute_error: 1.1995\n",
      "Epoch 91/200\n",
      "45/45 [==============================] - 4s 87ms/step - loss: 5.3449 - mean_absolute_error: 1.4494 - val_loss: 5.4672 - val_mean_absolute_error: 1.1253\n",
      "Epoch 92/200\n",
      "45/45 [==============================] - 3s 67ms/step - loss: 8.3709 - mean_absolute_error: 1.7284 - val_loss: 19.0139 - val_mean_absolute_error: 3.4720\n",
      "Epoch 93/200\n",
      "45/45 [==============================] - 3s 66ms/step - loss: 8.1956 - mean_absolute_error: 1.9496 - val_loss: 5.3024 - val_mean_absolute_error: 1.1803\n",
      "Epoch 94/200\n",
      "45/45 [==============================] - 3s 66ms/step - loss: 3.8523 - mean_absolute_error: 1.2577 - val_loss: 4.7510 - val_mean_absolute_error: 1.0440\n",
      "Epoch 95/200\n",
      "45/45 [==============================] - 4s 84ms/step - loss: 2.6771 - mean_absolute_error: 1.0731 - val_loss: 5.0613 - val_mean_absolute_error: 1.2092\n",
      "Epoch 96/200\n",
      "45/45 [==============================] - 3s 64ms/step - loss: 1.6680 - mean_absolute_error: 0.8600 - val_loss: 4.8959 - val_mean_absolute_error: 0.9862\n",
      "Epoch 97/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 7.6371 - mean_absolute_error: 1.6539 - val_loss: 7.4773 - val_mean_absolute_error: 1.4645\n",
      "Epoch 98/200\n",
      "45/45 [==============================] - 3s 66ms/step - loss: 3.7862 - mean_absolute_error: 1.3073 - val_loss: 4.6617 - val_mean_absolute_error: 1.0079\n",
      "Epoch 99/200\n",
      "45/45 [==============================] - 3s 74ms/step - loss: 2.4190 - mean_absolute_error: 0.9772 - val_loss: 4.8898 - val_mean_absolute_error: 1.2386\n",
      "Epoch 100/200\n",
      "45/45 [==============================] - 3s 67ms/step - loss: 11.4523 - mean_absolute_error: 2.3396 - val_loss: 8.2081 - val_mean_absolute_error: 1.6535\n",
      "Epoch 101/200\n",
      "45/45 [==============================] - 3s 69ms/step - loss: 20.7086 - mean_absolute_error: 3.0702 - val_loss: 12.4875 - val_mean_absolute_error: 2.6563\n",
      "Epoch 102/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 18.3554 - mean_absolute_error: 2.6415 - val_loss: 5.2449 - val_mean_absolute_error: 1.3231\n",
      "Epoch 103/200\n",
      "45/45 [==============================] - 3s 74ms/step - loss: 13.1885 - mean_absolute_error: 2.4339 - val_loss: 9.8926 - val_mean_absolute_error: 1.9574\n",
      "Epoch 104/200\n",
      "45/45 [==============================] - 3s 77ms/step - loss: 17.2643 - mean_absolute_error: 2.6567 - val_loss: 4.6731 - val_mean_absolute_error: 1.1389\n",
      "Epoch 105/200\n",
      "45/45 [==============================] - 3s 69ms/step - loss: 154.0130 - mean_absolute_error: 8.1916 - val_loss: 34.6688 - val_mean_absolute_error: 4.1478\n",
      "Epoch 106/200\n",
      "45/45 [==============================] - 3s 60ms/step - loss: 219.6590 - mean_absolute_error: 10.0357 - val_loss: 57.9752 - val_mean_absolute_error: 5.9107\n",
      "Epoch 107/200\n",
      "45/45 [==============================] - 3s 78ms/step - loss: 38.7537 - mean_absolute_error: 4.1112 - val_loss: 13.6134 - val_mean_absolute_error: 2.0577\n",
      "Epoch 108/200\n",
      "45/45 [==============================] - 4s 78ms/step - loss: 9.5610 - mean_absolute_error: 2.0790 - val_loss: 8.5490 - val_mean_absolute_error: 1.4752\n",
      "Epoch 109/200\n",
      "45/45 [==============================] - 3s 70ms/step - loss: 13.4150 - mean_absolute_error: 2.3669 - val_loss: 5.9181 - val_mean_absolute_error: 1.1554\n",
      "Epoch 110/200\n",
      "45/45 [==============================] - 3s 62ms/step - loss: 5.0485 - mean_absolute_error: 1.3938 - val_loss: 9.8962 - val_mean_absolute_error: 2.1060\n",
      "Epoch 111/200\n",
      "45/45 [==============================] - 3s 73ms/step - loss: 8.5833 - mean_absolute_error: 1.8313 - val_loss: 5.4901 - val_mean_absolute_error: 1.2204\n",
      "Epoch 112/200\n",
      "45/45 [==============================] - 3s 73ms/step - loss: 1.8684 - mean_absolute_error: 0.8633 - val_loss: 6.5943 - val_mean_absolute_error: 1.4548\n",
      "Epoch 113/200\n",
      "45/45 [==============================] - 3s 65ms/step - loss: 11.4097 - mean_absolute_error: 2.2642 - val_loss: 8.0718 - val_mean_absolute_error: 2.1045\n",
      "Epoch 114/200\n",
      "45/45 [==============================] - 3s 70ms/step - loss: 2.1130 - mean_absolute_error: 1.0206 - val_loss: 5.2027 - val_mean_absolute_error: 1.3277\n",
      "Epoch 115/200\n",
      "45/45 [==============================] - 3s 70ms/step - loss: 4.3866 - mean_absolute_error: 1.3426 - val_loss: 4.0057 - val_mean_absolute_error: 0.7401\n",
      "Epoch 116/200\n",
      "45/45 [==============================] - 3s 78ms/step - loss: 5.2722 - mean_absolute_error: 1.5280 - val_loss: 5.9726 - val_mean_absolute_error: 1.4337\n",
      "Epoch 117/200\n",
      "45/45 [==============================] - 3s 61ms/step - loss: 1.7760 - mean_absolute_error: 0.9529 - val_loss: 5.2563 - val_mean_absolute_error: 1.1850\n",
      "Epoch 118/200\n",
      "45/45 [==============================] - 3s 61ms/step - loss: 2.2435 - mean_absolute_error: 1.0326 - val_loss: 4.0997 - val_mean_absolute_error: 0.8238\n",
      "Epoch 119/200\n",
      "45/45 [==============================] - 3s 62ms/step - loss: 2.0698 - mean_absolute_error: 0.9950 - val_loss: 3.7152 - val_mean_absolute_error: 0.7740\n",
      "Epoch 120/200\n",
      "45/45 [==============================] - 4s 84ms/step - loss: 1.2101 - mean_absolute_error: 0.7594 - val_loss: 3.9406 - val_mean_absolute_error: 0.8933\n",
      "Epoch 121/200\n",
      "45/45 [==============================] - 3s 67ms/step - loss: 5.0545 - mean_absolute_error: 1.5120 - val_loss: 4.9892 - val_mean_absolute_error: 1.2205\n",
      "Epoch 122/200\n",
      "45/45 [==============================] - 3s 65ms/step - loss: 11.9663 - mean_absolute_error: 2.3485 - val_loss: 18.8931 - val_mean_absolute_error: 3.2477\n",
      "Epoch 123/200\n",
      "45/45 [==============================] - 3s 67ms/step - loss: 14.0369 - mean_absolute_error: 2.6011 - val_loss: 6.6118 - val_mean_absolute_error: 1.6604\n",
      "Epoch 124/200\n",
      "45/45 [==============================] - 3s 72ms/step - loss: 10.5295 - mean_absolute_error: 2.0940 - val_loss: 4.6270 - val_mean_absolute_error: 1.0546\n",
      "Epoch 125/200\n",
      "45/45 [==============================] - 3s 69ms/step - loss: 3.6105 - mean_absolute_error: 1.1720 - val_loss: 4.3903 - val_mean_absolute_error: 1.0287\n",
      "Epoch 126/200\n",
      "45/45 [==============================] - 2s 53ms/step - loss: 19.9233 - mean_absolute_error: 2.7352 - val_loss: 11.5545 - val_mean_absolute_error: 2.2959\n",
      "Epoch 127/200\n",
      "45/45 [==============================] - 3s 62ms/step - loss: 21.5366 - mean_absolute_error: 2.9863 - val_loss: 14.5676 - val_mean_absolute_error: 2.5083\n",
      "Epoch 128/200\n",
      "45/45 [==============================] - 2s 55ms/step - loss: 121.5490 - mean_absolute_error: 7.0196 - val_loss: 27.0929 - val_mean_absolute_error: 4.2632\n",
      "Epoch 129/200\n",
      "45/45 [==============================] - 3s 73ms/step - loss: 12.4305 - mean_absolute_error: 2.3929 - val_loss: 7.3430 - val_mean_absolute_error: 1.4699\n",
      "Epoch 130/200\n",
      "45/45 [==============================] - 3s 60ms/step - loss: 2.4199 - mean_absolute_error: 0.9900 - val_loss: 6.3352 - val_mean_absolute_error: 1.1400\n",
      "Epoch 131/200\n",
      "45/45 [==============================] - 3s 65ms/step - loss: 16.5441 - mean_absolute_error: 2.5771 - val_loss: 23.0234 - val_mean_absolute_error: 3.8495\n",
      "Epoch 132/200\n",
      "45/45 [==============================] - 3s 69ms/step - loss: 115.5461 - mean_absolute_error: 7.0646 - val_loss: 19.3566 - val_mean_absolute_error: 2.5496\n",
      "Epoch 133/200\n",
      "45/45 [==============================] - 4s 81ms/step - loss: 48.0544 - mean_absolute_error: 4.7873 - val_loss: 11.2906 - val_mean_absolute_error: 1.8321\n",
      "Epoch 134/200\n",
      "45/45 [==============================] - 3s 73ms/step - loss: 8.4568 - mean_absolute_error: 1.9064 - val_loss: 9.9013 - val_mean_absolute_error: 1.6268\n",
      "Epoch 135/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 16.0012 - mean_absolute_error: 2.4473 - val_loss: 8.1038 - val_mean_absolute_error: 2.0283\n",
      "Epoch 136/200\n",
      "45/45 [==============================] - 3s 66ms/step - loss: 4.7385 - mean_absolute_error: 1.4599 - val_loss: 5.7616 - val_mean_absolute_error: 1.2147\n",
      "Epoch 137/200\n",
      "45/45 [==============================] - 3s 69ms/step - loss: 3.4725 - mean_absolute_error: 1.1877 - val_loss: 4.7237 - val_mean_absolute_error: 1.0277\n",
      "Epoch 138/200\n",
      "45/45 [==============================] - 3s 76ms/step - loss: 3.8417 - mean_absolute_error: 1.2099 - val_loss: 5.6177 - val_mean_absolute_error: 1.2154\n",
      "Epoch 139/200\n",
      "45/45 [==============================] - 3s 69ms/step - loss: 15.6353 - mean_absolute_error: 2.3849 - val_loss: 17.2791 - val_mean_absolute_error: 3.3848\n",
      "Epoch 140/200\n",
      "45/45 [==============================] - 3s 60ms/step - loss: 10.5356 - mean_absolute_error: 2.1749 - val_loss: 6.3590 - val_mean_absolute_error: 1.1276\n",
      "Epoch 141/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 27.2948 - mean_absolute_error: 3.5183 - val_loss: 28.3806 - val_mean_absolute_error: 3.9602\n",
      "Epoch 142/200\n",
      "45/45 [==============================] - 3s 72ms/step - loss: 13.0006 - mean_absolute_error: 2.3896 - val_loss: 8.3673 - val_mean_absolute_error: 1.7162\n",
      "Epoch 143/200\n",
      "45/45 [==============================] - 3s 67ms/step - loss: 5.7271 - mean_absolute_error: 1.6093 - val_loss: 6.5529 - val_mean_absolute_error: 1.4358\n",
      "Epoch 144/200\n",
      "45/45 [==============================] - 3s 60ms/step - loss: 9.2482 - mean_absolute_error: 2.0256 - val_loss: 5.5645 - val_mean_absolute_error: 1.1551\n",
      "Epoch 145/200\n",
      "45/45 [==============================] - 3s 65ms/step - loss: 14.2159 - mean_absolute_error: 2.4332 - val_loss: 8.9539 - val_mean_absolute_error: 1.8298\n",
      "Epoch 146/200\n",
      "45/45 [==============================] - 4s 82ms/step - loss: 4.9713 - mean_absolute_error: 1.4438 - val_loss: 11.2002 - val_mean_absolute_error: 2.3846\n",
      "Epoch 147/200\n",
      "45/45 [==============================] - 3s 63ms/step - loss: 119.2755 - mean_absolute_error: 7.0172 - val_loss: 107.5014 - val_mean_absolute_error: 8.0791\n",
      "Epoch 148/200\n",
      "45/45 [==============================] - 3s 60ms/step - loss: 73.8233 - mean_absolute_error: 6.1038 - val_loss: 28.0586 - val_mean_absolute_error: 3.7543\n",
      "Epoch 149/200\n",
      "45/45 [==============================] - 2s 54ms/step - loss: 46.5549 - mean_absolute_error: 4.5343 - val_loss: 12.9262 - val_mean_absolute_error: 2.0073\n",
      "Epoch 150/200\n",
      "45/45 [==============================] - 3s 71ms/step - loss: 50.7887 - mean_absolute_error: 4.4861 - val_loss: 28.2386 - val_mean_absolute_error: 3.8705\n",
      "Epoch 151/200\n",
      "45/45 [==============================] - 3s 74ms/step - loss: 7.5816 - mean_absolute_error: 1.7982 - val_loss: 6.6631 - val_mean_absolute_error: 1.3562\n",
      "Epoch 152/200\n",
      "45/45 [==============================] - 3s 59ms/step - loss: 7.1275 - mean_absolute_error: 1.6593 - val_loss: 5.4715 - val_mean_absolute_error: 1.1869\n",
      "Epoch 153/200\n",
      "45/45 [==============================] - 3s 65ms/step - loss: 2.0598 - mean_absolute_error: 0.9614 - val_loss: 5.1855 - val_mean_absolute_error: 1.0910\n",
      "Epoch 154/200\n",
      "45/45 [==============================] - 3s 58ms/step - loss: 12.4339 - mean_absolute_error: 2.0996 - val_loss: 17.7981 - val_mean_absolute_error: 3.2978\n",
      "Epoch 155/200\n",
      "45/45 [==============================] - 3s 72ms/step - loss: 21.6520 - mean_absolute_error: 3.1557 - val_loss: 35.8035 - val_mean_absolute_error: 4.4824\n",
      "Epoch 156/200\n",
      "45/45 [==============================] - 3s 61ms/step - loss: 30.5674 - mean_absolute_error: 3.3737 - val_loss: 11.4766 - val_mean_absolute_error: 2.2028\n",
      "Epoch 157/200\n",
      "45/45 [==============================] - 3s 61ms/step - loss: 18.1449 - mean_absolute_error: 2.9416 - val_loss: 8.3847 - val_mean_absolute_error: 1.9585\n",
      "Epoch 158/200\n",
      "45/45 [==============================] - 3s 59ms/step - loss: 7.3490 - mean_absolute_error: 1.7895 - val_loss: 5.1059 - val_mean_absolute_error: 1.1907\n",
      "Epoch 159/200\n",
      "45/45 [==============================] - 3s 69ms/step - loss: 3.4582 - mean_absolute_error: 1.1936 - val_loss: 4.0158 - val_mean_absolute_error: 0.9170\n",
      "Epoch 160/200\n",
      "45/45 [==============================] - 3s 69ms/step - loss: 7.2011 - mean_absolute_error: 1.7807 - val_loss: 7.1287 - val_mean_absolute_error: 1.7180\n",
      "Epoch 161/200\n",
      "45/45 [==============================] - 2s 54ms/step - loss: 7.2673 - mean_absolute_error: 1.8499 - val_loss: 5.4744 - val_mean_absolute_error: 1.2080\n",
      "Epoch 162/200\n",
      "45/45 [==============================] - 3s 65ms/step - loss: 3.5221 - mean_absolute_error: 1.2338 - val_loss: 3.7630 - val_mean_absolute_error: 0.8763\n",
      "Epoch 163/200\n",
      "45/45 [==============================] - 3s 61ms/step - loss: 2.3795 - mean_absolute_error: 0.9629 - val_loss: 4.7859 - val_mean_absolute_error: 1.1557\n",
      "Epoch 164/200\n",
      "45/45 [==============================] - 4s 80ms/step - loss: 8.3236 - mean_absolute_error: 1.9589 - val_loss: 5.0108 - val_mean_absolute_error: 1.3121\n",
      "Epoch 165/200\n",
      "45/45 [==============================] - 3s 75ms/step - loss: 8.2759 - mean_absolute_error: 1.9293 - val_loss: 5.7580 - val_mean_absolute_error: 1.5208\n",
      "Epoch 166/200\n",
      "45/45 [==============================] - 3s 66ms/step - loss: 4.2133 - mean_absolute_error: 1.2928 - val_loss: 5.9453 - val_mean_absolute_error: 1.6304\n",
      "Epoch 167/200\n",
      "45/45 [==============================] - 3s 60ms/step - loss: 92.5571 - mean_absolute_error: 5.3903 - val_loss: 29.4276 - val_mean_absolute_error: 4.2616\n",
      "Epoch 168/200\n",
      "45/45 [==============================] - 3s 69ms/step - loss: 49.4544 - mean_absolute_error: 4.6401 - val_loss: 13.6624 - val_mean_absolute_error: 2.4191\n",
      "Epoch 169/200\n",
      "45/45 [==============================] - 3s 63ms/step - loss: 7.7001 - mean_absolute_error: 1.8770 - val_loss: 6.4641 - val_mean_absolute_error: 1.3150\n",
      "Epoch 170/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 3.8319 - mean_absolute_error: 1.2853 - val_loss: 5.5145 - val_mean_absolute_error: 1.4021\n",
      "Epoch 171/200\n",
      "45/45 [==============================] - 3s 71ms/step - loss: 2.3506 - mean_absolute_error: 0.9552 - val_loss: 4.0388 - val_mean_absolute_error: 1.0521\n",
      "Epoch 172/200\n",
      "45/45 [==============================] - 3s 73ms/step - loss: 2.8594 - mean_absolute_error: 1.1135 - val_loss: 5.3464 - val_mean_absolute_error: 1.3364\n",
      "Epoch 173/200\n",
      "45/45 [==============================] - 4s 81ms/step - loss: 15.4077 - mean_absolute_error: 2.5894 - val_loss: 11.5442 - val_mean_absolute_error: 2.3211\n",
      "Epoch 174/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 69.3278 - mean_absolute_error: 5.0422 - val_loss: 27.3789 - val_mean_absolute_error: 2.6629\n",
      "Epoch 175/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 1109.8673 - mean_absolute_error: 21.5076 - val_loss: 238.4135 - val_mean_absolute_error: 12.5447\n",
      "Epoch 176/200\n",
      "45/45 [==============================] - 3s 72ms/step - loss: 162.5532 - mean_absolute_error: 7.3387 - val_loss: 28.7818 - val_mean_absolute_error: 2.6058\n",
      "Epoch 177/200\n",
      "45/45 [==============================] - 4s 96ms/step - loss: 22.1034 - mean_absolute_error: 3.0735 - val_loss: 19.7843 - val_mean_absolute_error: 2.4082\n",
      "Epoch 178/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 22.4616 - mean_absolute_error: 3.0369 - val_loss: 12.7399 - val_mean_absolute_error: 1.6713\n",
      "Epoch 179/200\n",
      "45/45 [==============================] - 3s 60ms/step - loss: 4.1863 - mean_absolute_error: 1.2858 - val_loss: 10.8526 - val_mean_absolute_error: 1.6198\n",
      "Epoch 180/200\n",
      "45/45 [==============================] - 3s 55ms/step - loss: 3.9043 - mean_absolute_error: 1.2640 - val_loss: 8.8023 - val_mean_absolute_error: 1.2252\n",
      "Epoch 181/200\n",
      "45/45 [==============================] - 3s 75ms/step - loss: 4.3553 - mean_absolute_error: 1.2840 - val_loss: 7.5498 - val_mean_absolute_error: 1.0698\n",
      "Epoch 182/200\n",
      "45/45 [==============================] - 3s 71ms/step - loss: 1.3854 - mean_absolute_error: 0.7315 - val_loss: 6.8914 - val_mean_absolute_error: 0.9776\n",
      "Epoch 183/200\n",
      "45/45 [==============================] - 3s 58ms/step - loss: 2.8989 - mean_absolute_error: 1.1060 - val_loss: 7.2306 - val_mean_absolute_error: 1.1330\n",
      "Epoch 184/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 5.6086 - mean_absolute_error: 1.5128 - val_loss: 7.0364 - val_mean_absolute_error: 1.3925\n",
      "Epoch 185/200\n",
      "45/45 [==============================] - 2s 54ms/step - loss: 1.4281 - mean_absolute_error: 0.8104 - val_loss: 6.4078 - val_mean_absolute_error: 1.0020\n",
      "Epoch 186/200\n",
      "45/45 [==============================] - 3s 68ms/step - loss: 0.8709 - mean_absolute_error: 0.6220 - val_loss: 5.8786 - val_mean_absolute_error: 1.0039\n",
      "Epoch 187/200\n",
      "45/45 [==============================] - 3s 65ms/step - loss: 1.0902 - mean_absolute_error: 0.6669 - val_loss: 5.5005 - val_mean_absolute_error: 0.8166\n",
      "Epoch 188/200\n",
      "45/45 [==============================] - 3s 60ms/step - loss: 0.8820 - mean_absolute_error: 0.6345 - val_loss: 5.4783 - val_mean_absolute_error: 0.7745\n",
      "Epoch 189/200\n",
      "45/45 [==============================] - 3s 58ms/step - loss: 0.6546 - mean_absolute_error: 0.5317 - val_loss: 5.1774 - val_mean_absolute_error: 0.7995\n",
      "Epoch 190/200\n",
      "45/45 [==============================] - 3s 61ms/step - loss: 0.6309 - mean_absolute_error: 0.5202 - val_loss: 5.1045 - val_mean_absolute_error: 0.7444\n",
      "Epoch 191/200\n",
      "45/45 [==============================] - 3s 76ms/step - loss: 0.6095 - mean_absolute_error: 0.5228 - val_loss: 5.0674 - val_mean_absolute_error: 0.8238\n",
      "Epoch 192/200\n",
      "45/45 [==============================] - 3s 61ms/step - loss: 5.4919 - mean_absolute_error: 1.4070 - val_loss: 5.6017 - val_mean_absolute_error: 1.1100\n",
      "Epoch 193/200\n",
      "45/45 [==============================] - 3s 59ms/step - loss: 3.7044 - mean_absolute_error: 1.1692 - val_loss: 4.5315 - val_mean_absolute_error: 0.7273\n",
      "Epoch 194/200\n",
      "45/45 [==============================] - 3s 62ms/step - loss: 0.3470 - mean_absolute_error: 0.3889 - val_loss: 4.2199 - val_mean_absolute_error: 0.6370\n",
      "Epoch 195/200\n",
      "45/45 [==============================] - 3s 75ms/step - loss: 0.3520 - mean_absolute_error: 0.3937 - val_loss: 4.1359 - val_mean_absolute_error: 0.6533\n",
      "Epoch 196/200\n",
      "45/45 [==============================] - 3s 63ms/step - loss: 0.2953 - mean_absolute_error: 0.3656 - val_loss: 4.1875 - val_mean_absolute_error: 0.6403\n",
      "Epoch 197/200\n",
      "45/45 [==============================] - 3s 56ms/step - loss: 1.1026 - mean_absolute_error: 0.6477 - val_loss: 4.3298 - val_mean_absolute_error: 0.7084\n",
      "Epoch 198/200\n",
      "45/45 [==============================] - 3s 61ms/step - loss: 1.0123 - mean_absolute_error: 0.6447 - val_loss: 4.4204 - val_mean_absolute_error: 1.0086\n",
      "Epoch 199/200\n",
      "45/45 [==============================] - 2s 53ms/step - loss: 0.9872 - mean_absolute_error: 0.6492 - val_loss: 3.7998 - val_mean_absolute_error: 0.5417\n",
      "Epoch 200/200\n",
      "45/45 [==============================] - 4s 86ms/step - loss: 0.4529 - mean_absolute_error: 0.4506 - val_loss: 3.9493 - val_mean_absolute_error: 0.8300\n"
     ]
    }
   ],
   "source": [
    "# Define a model with increased complexity\n",
    "overfitting_model = tf.keras.Sequential([\n",
    "    tf.keras.layers.Dense(512, activation='relu', input_shape=(X_train.shape[1],)),\n",
    "    tf.keras.layers.Dense(1024, activation='relu'),\n",
    "    tf.keras.layers.Dense(2048, activation='relu'),\n",
    "    tf.keras.layers.Dense(1)\n",
    "])\n",
    "\n",
    "# Compile the model without regularization\n",
    "overfitting_model.compile(loss='mean_squared_error', optimizer='adam', metrics=['mean_absolute_error'])\n",
    "\n",
    "# Train the model for more epochs\n",
    "overfitting_history = overfitting_model.fit(X_train_normalized, y_train, epochs=200, batch_size=32, validation_split=0.2)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 607
    },
    "id": "533w6Yqp4-1z",
    "outputId": "108589d2-7dab-4d7f-de01-693dafa571df"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot\n",
    "fig, axs = plt.subplots(1, 2, figsize=(12, 6))\n",
    "\n",
    "# training & validation loss values plot\n",
    "axs[0].plot(overfitting_history.history['loss'], label='Training Loss')\n",
    "axs[0].plot(overfitting_history.history['val_loss'], label='Validation Loss')\n",
    "axs[0].set_title('Model Loss')\n",
    "axs[0].set_xlabel('Epoch')\n",
    "axs[0].set_ylabel('Loss')\n",
    "axs[0].legend(['Train', 'Validation'], loc='upper right')\n",
    "\n",
    "# training & validation mean absolute error values plot\n",
    "axs[1].plot(overfitting_history.history['mean_absolute_error'], label='Training MAE')\n",
    "axs[1].plot(overfitting_history.history['val_mean_absolute_error'], label='Validation MAE')\n",
    "axs[1].set_title('Mean Absolute Error')\n",
    "axs[1].set_xlabel('Epoch')\n",
    "axs[1].set_ylabel('MAE')\n",
    "axs[1].legend(['Train', 'Validation'], loc='upper right')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "JDXYxvw0AXUx"
   },
   "source": [
    "If the model loss quickly decreases to near-zero values on both the training and validation datasets, it is likely that the model is overfitting rather than being a sign of good performance."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "sosUArG3_x86"
   },
   "source": [
    "<h2>7. Regularizing model and tuning hyperparameters</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "BbpXIUWWAew8"
   },
   "source": [
    "Tuning hyperparameters with Keras Tuner"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "id": "3orKY9tp_zOS"
   },
   "outputs": [],
   "source": [
    "#!pip install keras-tuner --upgrade\n",
    "\n",
    "import keras_tuner\n",
    "\n",
    "from kerastuner.tuners import RandomSearch\n",
    "from keras.layers import Dense\n",
    "\n",
    "# define a hyperparameter tuning function\n",
    "def build_model(hp):\n",
    "    model = tf.keras.Sequential()\n",
    "    model.add(Dense(units=hp.Int('units1', min_value=64, max_value=512, step=64), activation='relu', input_shape=(X_train_normalized.shape[1],)))\n",
    "\n",
    "    # add more hidden layers based on hyperparameters\n",
    "    for i in range(hp.Int('num_layers', min_value=1, max_value=4)):\n",
    "        model.add(Dense(units=hp.Int(f'units_{i}', min_value=64, max_value=512, step=64), activation='relu'))\n",
    "\n",
    "    model.add(Dense(1))\n",
    "\n",
    "    model.compile(loss='mean_squared_error', optimizer='adam', metrics=['mean_absolute_error'])\n",
    "\n",
    "    return model\n",
    "\n",
    "# create Keras Tuner RandomSearch tuner\n",
    "tuner = RandomSearch(\n",
    "    build_model,\n",
    "    objective='val_loss',\n",
    "    max_trials=5,  # number of hyperparameter combinations to try\n",
    "    executions_per_trial=1,\n",
    "    directory='my_tuner_dir'\n",
    ")\n",
    "\n",
    "# perform hyperparameter tuning\n",
    "tuner.search(X_train_normalized, y_train, epochs=100, batch_size=32, validation_split=0.2)\n",
    "\n",
    "# get the best hyperparameters\n",
    "best_hp = tuner.get_best_hyperparameters(1)[0]\n",
    "best_model = tuner.hypermodel.build(best_hp)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "nRK9RvoJhUj7"
   },
   "source": [
    "View results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "OzkYVDz-Ag7x",
    "outputId": "3b6ec10a-d732-4ddf-b4c5-b50bbe7efb1d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Results summary\n",
      "Results in my_tuner_dir/untitled_project\n",
      "Showing 10 best trials\n",
      "Objective(name=\"val_loss\", direction=\"min\")\n",
      "\n",
      "Trial 0 summary\n",
      "Hyperparameters:\n",
      "units1: 320\n",
      "num_layers: 1\n",
      "units_0: 512\n",
      "Score: 9400.8935546875\n",
      "\n",
      "Trial 2 summary\n",
      "Hyperparameters:\n",
      "units1: 64\n",
      "num_layers: 2\n",
      "units_0: 192\n",
      "units_1: 64\n",
      "Score: 9840.3291015625\n",
      "\n",
      "Trial 1 summary\n",
      "Hyperparameters:\n",
      "units1: 192\n",
      "num_layers: 1\n",
      "units_0: 320\n",
      "Score: 10146.1689453125\n",
      "\n",
      "Trial 4 summary\n",
      "Hyperparameters:\n",
      "units1: 448\n",
      "num_layers: 3\n",
      "units_0: 448\n",
      "units_1: 320\n",
      "units_2: 256\n",
      "Score: 33032.74609375\n",
      "\n",
      "Trial 3 summary\n",
      "Hyperparameters:\n",
      "units1: 384\n",
      "num_layers: 3\n",
      "units_0: 128\n",
      "units_1: 320\n",
      "units_2: 64\n",
      "Score: 46593.25390625\n",
      "Epoch 1/50\n",
      "45/45 [==============================] - 0s 8ms/step - loss: 2.0013 - mean_absolute_error: 0.8778 - val_loss: 7.3827 - val_mean_absolute_error: 1.3326\n",
      "Epoch 2/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 2.0453 - mean_absolute_error: 0.9401 - val_loss: 8.9833 - val_mean_absolute_error: 1.9012\n",
      "Epoch 3/50\n",
      "45/45 [==============================] - 0s 8ms/step - loss: 2.1297 - mean_absolute_error: 0.9744 - val_loss: 6.1918 - val_mean_absolute_error: 0.9600\n",
      "Epoch 4/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 2.0032 - mean_absolute_error: 0.9230 - val_loss: 5.8995 - val_mean_absolute_error: 0.8669\n",
      "Epoch 5/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 1.9193 - mean_absolute_error: 0.9481 - val_loss: 5.7390 - val_mean_absolute_error: 0.8744\n",
      "Epoch 6/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 1.7646 - mean_absolute_error: 0.8450 - val_loss: 5.7239 - val_mean_absolute_error: 0.9404\n",
      "Epoch 7/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 2.1307 - mean_absolute_error: 1.0114 - val_loss: 5.7964 - val_mean_absolute_error: 0.8998\n",
      "Epoch 8/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 1.8616 - mean_absolute_error: 0.9201 - val_loss: 5.6114 - val_mean_absolute_error: 1.0778\n",
      "Epoch 9/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 1.6911 - mean_absolute_error: 0.8520 - val_loss: 5.4345 - val_mean_absolute_error: 0.8127\n",
      "Epoch 10/50\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 1.3375 - mean_absolute_error: 0.7402 - val_loss: 5.9017 - val_mean_absolute_error: 1.1027\n",
      "Epoch 11/50\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 1.5014 - mean_absolute_error: 0.7995 - val_loss: 5.9442 - val_mean_absolute_error: 1.0941\n",
      "Epoch 12/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 1.4972 - mean_absolute_error: 0.8019 - val_loss: 5.2111 - val_mean_absolute_error: 0.7989\n",
      "Epoch 13/50\n",
      "45/45 [==============================] - 0s 8ms/step - loss: 1.7992 - mean_absolute_error: 0.9259 - val_loss: 7.3299 - val_mean_absolute_error: 1.5243\n",
      "Epoch 14/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 1.7918 - mean_absolute_error: 0.9687 - val_loss: 7.7394 - val_mean_absolute_error: 1.7117\n",
      "Epoch 15/50\n",
      "45/45 [==============================] - 0s 8ms/step - loss: 1.6013 - mean_absolute_error: 0.8354 - val_loss: 4.8351 - val_mean_absolute_error: 0.7794\n",
      "Epoch 16/50\n",
      "45/45 [==============================] - 0s 8ms/step - loss: 1.2271 - mean_absolute_error: 0.7263 - val_loss: 4.4805 - val_mean_absolute_error: 0.7509\n",
      "Epoch 17/50\n",
      "45/45 [==============================] - 0s 8ms/step - loss: 1.2961 - mean_absolute_error: 0.7850 - val_loss: 4.4180 - val_mean_absolute_error: 0.6576\n",
      "Epoch 18/50\n",
      "45/45 [==============================] - 0s 9ms/step - loss: 1.4915 - mean_absolute_error: 0.8536 - val_loss: 4.3893 - val_mean_absolute_error: 0.6831\n",
      "Epoch 19/50\n",
      "45/45 [==============================] - 0s 8ms/step - loss: 1.1957 - mean_absolute_error: 0.7380 - val_loss: 4.3340 - val_mean_absolute_error: 0.6497\n",
      "Epoch 20/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 1.1763 - mean_absolute_error: 0.7278 - val_loss: 4.2729 - val_mean_absolute_error: 0.7174\n",
      "Epoch 21/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 0.9555 - mean_absolute_error: 0.6769 - val_loss: 4.1640 - val_mean_absolute_error: 0.6368\n",
      "Epoch 22/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 1.3608 - mean_absolute_error: 0.8247 - val_loss: 6.1200 - val_mean_absolute_error: 1.4817\n",
      "Epoch 23/50\n",
      "45/45 [==============================] - 1s 12ms/step - loss: 0.9877 - mean_absolute_error: 0.6809 - val_loss: 4.2192 - val_mean_absolute_error: 0.7337\n",
      "Epoch 24/50\n",
      "45/45 [==============================] - 0s 10ms/step - loss: 0.7723 - mean_absolute_error: 0.5741 - val_loss: 5.6000 - val_mean_absolute_error: 1.3186\n",
      "Epoch 25/50\n",
      "45/45 [==============================] - 1s 11ms/step - loss: 1.0493 - mean_absolute_error: 0.7177 - val_loss: 4.4713 - val_mean_absolute_error: 0.8380\n",
      "Epoch 26/50\n",
      "45/45 [==============================] - 1s 12ms/step - loss: 1.3538 - mean_absolute_error: 0.8131 - val_loss: 3.9845 - val_mean_absolute_error: 0.6877\n",
      "Epoch 27/50\n",
      "45/45 [==============================] - 0s 10ms/step - loss: 0.9598 - mean_absolute_error: 0.6724 - val_loss: 4.7641 - val_mean_absolute_error: 1.1249\n",
      "Epoch 28/50\n",
      "45/45 [==============================] - 1s 12ms/step - loss: 0.8641 - mean_absolute_error: 0.6458 - val_loss: 4.2141 - val_mean_absolute_error: 0.7130\n",
      "Epoch 29/50\n",
      "45/45 [==============================] - 0s 9ms/step - loss: 1.0655 - mean_absolute_error: 0.7274 - val_loss: 4.1020 - val_mean_absolute_error: 0.7277\n",
      "Epoch 30/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 0.9406 - mean_absolute_error: 0.6724 - val_loss: 4.3197 - val_mean_absolute_error: 0.8726\n",
      "Epoch 31/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 0.7590 - mean_absolute_error: 0.5992 - val_loss: 3.7253 - val_mean_absolute_error: 0.5540\n",
      "Epoch 32/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 0.5941 - mean_absolute_error: 0.5200 - val_loss: 3.6722 - val_mean_absolute_error: 0.4797\n",
      "Epoch 33/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 1.2626 - mean_absolute_error: 0.7979 - val_loss: 3.7373 - val_mean_absolute_error: 0.5762\n",
      "Epoch 34/50\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 0.6391 - mean_absolute_error: 0.5418 - val_loss: 3.6417 - val_mean_absolute_error: 0.5650\n",
      "Epoch 35/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 1.0627 - mean_absolute_error: 0.7413 - val_loss: 3.7778 - val_mean_absolute_error: 0.6896\n",
      "Epoch 36/50\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 1.2384 - mean_absolute_error: 0.7941 - val_loss: 3.7986 - val_mean_absolute_error: 0.7790\n",
      "Epoch 37/50\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 0.6363 - mean_absolute_error: 0.5650 - val_loss: 3.6076 - val_mean_absolute_error: 0.5850\n",
      "Epoch 38/50\n",
      "45/45 [==============================] - 0s 5ms/step - loss: 0.4667 - mean_absolute_error: 0.4794 - val_loss: 3.6724 - val_mean_absolute_error: 0.6786\n",
      "Epoch 39/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 0.8896 - mean_absolute_error: 0.6859 - val_loss: 3.4334 - val_mean_absolute_error: 0.5736\n",
      "Epoch 40/50\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 0.9420 - mean_absolute_error: 0.6994 - val_loss: 3.4210 - val_mean_absolute_error: 0.6145\n",
      "Epoch 41/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 1.5913 - mean_absolute_error: 0.9159 - val_loss: 3.6692 - val_mean_absolute_error: 0.7226\n",
      "Epoch 42/50\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 1.0390 - mean_absolute_error: 0.7568 - val_loss: 3.2778 - val_mean_absolute_error: 0.6553\n",
      "Epoch 43/50\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 0.6653 - mean_absolute_error: 0.5848 - val_loss: 5.1974 - val_mean_absolute_error: 1.3675\n",
      "Epoch 44/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 1.0746 - mean_absolute_error: 0.7478 - val_loss: 4.0262 - val_mean_absolute_error: 1.0405\n",
      "Epoch 45/50\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 0.5982 - mean_absolute_error: 0.5554 - val_loss: 3.6164 - val_mean_absolute_error: 0.7458\n",
      "Epoch 46/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 0.8576 - mean_absolute_error: 0.6957 - val_loss: 3.1904 - val_mean_absolute_error: 0.4804\n",
      "Epoch 47/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 0.5360 - mean_absolute_error: 0.5377 - val_loss: 3.5364 - val_mean_absolute_error: 0.7979\n",
      "Epoch 48/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 0.5855 - mean_absolute_error: 0.5463 - val_loss: 6.3257 - val_mean_absolute_error: 1.7122\n",
      "Epoch 49/50\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 1.1255 - mean_absolute_error: 0.7993 - val_loss: 3.2953 - val_mean_absolute_error: 0.6244\n",
      "Epoch 50/50\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 1.4431 - mean_absolute_error: 0.8761 - val_loss: 3.3130 - val_mean_absolute_error: 0.6614\n"
     ]
    }
   ],
   "source": [
    "tuner.results_summary()\n",
    "\n",
    "trials = tuner.oracle.get_best_trials(num_trials=5)\n",
    "\n",
    "# extract hyperparameters and validation losses\n",
    "hyperparameters = [trial.hyperparameters.values for trial in trials]\n",
    "validation_losses = [-trial.score for trial in trials]  # negative score for validation loss\n",
    "\n",
    "# extract trial numbers for x-axis\n",
    "trial_numbers = list(range(1, len(hyperparameters) + 1))\n",
    "\n",
    "history = best_model.fit(X_train_normalized, y_train, epochs=50, batch_size=32, validation_split=0.2)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "aKPle0UxgG3S",
    "outputId": "1ed3fde7-3256-4e6d-f8f0-25f3eb759755"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "45/45 [==============================] - 2s 22ms/step - loss: 212049.7031 - mean_absolute_error: 299.6267 - val_loss: 8379.7197 - val_mean_absolute_error: 33.1125\n",
      "Epoch 2/50\n",
      "45/45 [==============================] - 1s 15ms/step - loss: 110510.4609 - mean_absolute_error: 204.8559 - val_loss: 4269.5928 - val_mean_absolute_error: 48.0885\n",
      "Epoch 3/50\n",
      "45/45 [==============================] - 1s 12ms/step - loss: 26043.4102 - mean_absolute_error: 111.1170 - val_loss: 5839.4146 - val_mean_absolute_error: 47.4875\n",
      "Epoch 4/50\n",
      "45/45 [==============================] - 1s 14ms/step - loss: 15906.7002 - mean_absolute_error: 90.2234 - val_loss: 4665.2710 - val_mean_absolute_error: 49.3913\n",
      "Epoch 5/50\n",
      "45/45 [==============================] - 1s 14ms/step - loss: 11267.0342 - mean_absolute_error: 79.4798 - val_loss: 3594.3760 - val_mean_absolute_error: 48.3653\n",
      "Epoch 6/50\n",
      "45/45 [==============================] - 0s 10ms/step - loss: 8556.9424 - mean_absolute_error: 70.9249 - val_loss: 2496.0623 - val_mean_absolute_error: 40.9442\n",
      "Epoch 7/50\n",
      "45/45 [==============================] - 0s 10ms/step - loss: 6993.8770 - mean_absolute_error: 63.5403 - val_loss: 1945.4889 - val_mean_absolute_error: 34.2750\n",
      "Epoch 8/50\n",
      "45/45 [==============================] - 1s 12ms/step - loss: 5915.3906 - mean_absolute_error: 56.1670 - val_loss: 1687.8379 - val_mean_absolute_error: 30.7781\n",
      "Epoch 9/50\n",
      "45/45 [==============================] - 1s 11ms/step - loss: 5009.1729 - mean_absolute_error: 51.9442 - val_loss: 1339.1704 - val_mean_absolute_error: 26.2382\n",
      "Epoch 10/50\n",
      "45/45 [==============================] - 1s 16ms/step - loss: 4180.9233 - mean_absolute_error: 45.7593 - val_loss: 1028.1750 - val_mean_absolute_error: 21.6192\n",
      "Epoch 11/50\n",
      "45/45 [==============================] - 1s 14ms/step - loss: 3373.1406 - mean_absolute_error: 39.8313 - val_loss: 866.8815 - val_mean_absolute_error: 19.7231\n",
      "Epoch 12/50\n",
      "45/45 [==============================] - 1s 21ms/step - loss: 2665.0447 - mean_absolute_error: 35.1026 - val_loss: 703.6545 - val_mean_absolute_error: 18.1909\n",
      "Epoch 13/50\n",
      "45/45 [==============================] - 1s 17ms/step - loss: 1950.3918 - mean_absolute_error: 29.7918 - val_loss: 568.5370 - val_mean_absolute_error: 17.6052\n",
      "Epoch 14/50\n",
      "45/45 [==============================] - 1s 18ms/step - loss: 1449.9305 - mean_absolute_error: 25.5028 - val_loss: 458.0179 - val_mean_absolute_error: 16.4247\n",
      "Epoch 15/50\n",
      "45/45 [==============================] - 1s 13ms/step - loss: 1131.6498 - mean_absolute_error: 23.0561 - val_loss: 467.5754 - val_mean_absolute_error: 17.6836\n",
      "Epoch 16/50\n",
      "45/45 [==============================] - 1s 12ms/step - loss: 913.6984 - mean_absolute_error: 20.6609 - val_loss: 367.4744 - val_mean_absolute_error: 14.8631\n",
      "Epoch 17/50\n",
      "45/45 [==============================] - 1s 11ms/step - loss: 794.6906 - mean_absolute_error: 19.1575 - val_loss: 336.4789 - val_mean_absolute_error: 14.4785\n",
      "Epoch 18/50\n",
      "45/45 [==============================] - 1s 16ms/step - loss: 679.7068 - mean_absolute_error: 17.9478 - val_loss: 330.1221 - val_mean_absolute_error: 14.8252\n",
      "Epoch 19/50\n",
      "45/45 [==============================] - 1s 15ms/step - loss: 605.9744 - mean_absolute_error: 16.5309 - val_loss: 250.9198 - val_mean_absolute_error: 12.0719\n",
      "Epoch 20/50\n",
      "45/45 [==============================] - 1s 12ms/step - loss: 533.3782 - mean_absolute_error: 15.7006 - val_loss: 200.8731 - val_mean_absolute_error: 10.1148\n",
      "Epoch 21/50\n",
      "45/45 [==============================] - 1s 12ms/step - loss: 477.3597 - mean_absolute_error: 14.5666 - val_loss: 245.1555 - val_mean_absolute_error: 12.1557\n",
      "Epoch 22/50\n",
      "45/45 [==============================] - 0s 11ms/step - loss: 427.3987 - mean_absolute_error: 13.6940 - val_loss: 153.0178 - val_mean_absolute_error: 8.3801\n",
      "Epoch 23/50\n",
      "45/45 [==============================] - 0s 11ms/step - loss: 378.4294 - mean_absolute_error: 12.6429 - val_loss: 192.3497 - val_mean_absolute_error: 10.3770\n",
      "Epoch 24/50\n",
      "45/45 [==============================] - 1s 12ms/step - loss: 341.8500 - mean_absolute_error: 11.8629 - val_loss: 147.3358 - val_mean_absolute_error: 8.3873\n",
      "Epoch 25/50\n",
      "45/45 [==============================] - 1s 14ms/step - loss: 305.2513 - mean_absolute_error: 10.9991 - val_loss: 145.5582 - val_mean_absolute_error: 8.1652\n",
      "Epoch 26/50\n",
      "45/45 [==============================] - 1s 17ms/step - loss: 274.6168 - mean_absolute_error: 10.4716 - val_loss: 118.6159 - val_mean_absolute_error: 7.0510\n",
      "Epoch 27/50\n",
      "45/45 [==============================] - 1s 15ms/step - loss: 259.2975 - mean_absolute_error: 10.2263 - val_loss: 106.6710 - val_mean_absolute_error: 6.3277\n",
      "Epoch 28/50\n",
      "45/45 [==============================] - 1s 14ms/step - loss: 225.6991 - mean_absolute_error: 9.3038 - val_loss: 116.7225 - val_mean_absolute_error: 7.0447\n",
      "Epoch 29/50\n",
      "45/45 [==============================] - 1s 15ms/step - loss: 197.4570 - mean_absolute_error: 8.2894 - val_loss: 117.0338 - val_mean_absolute_error: 7.1563\n",
      "Epoch 30/50\n",
      "45/45 [==============================] - 1s 17ms/step - loss: 183.6647 - mean_absolute_error: 7.9374 - val_loss: 106.3324 - val_mean_absolute_error: 6.5147\n",
      "Epoch 31/50\n",
      "45/45 [==============================] - 1s 21ms/step - loss: 169.1480 - mean_absolute_error: 7.8312 - val_loss: 115.5901 - val_mean_absolute_error: 7.0735\n",
      "Epoch 32/50\n",
      "45/45 [==============================] - 1s 15ms/step - loss: 149.3783 - mean_absolute_error: 7.0612 - val_loss: 104.8912 - val_mean_absolute_error: 6.6446\n",
      "Epoch 33/50\n",
      "45/45 [==============================] - 1s 13ms/step - loss: 139.0957 - mean_absolute_error: 6.6269 - val_loss: 115.0147 - val_mean_absolute_error: 7.1102\n",
      "Epoch 34/50\n",
      "45/45 [==============================] - 1s 18ms/step - loss: 125.7873 - mean_absolute_error: 6.2890 - val_loss: 76.6539 - val_mean_absolute_error: 5.1624\n",
      "Epoch 35/50\n",
      "45/45 [==============================] - 1s 14ms/step - loss: 118.3378 - mean_absolute_error: 6.2988 - val_loss: 70.0433 - val_mean_absolute_error: 4.6272\n",
      "Epoch 36/50\n",
      "45/45 [==============================] - 1s 13ms/step - loss: 100.2744 - mean_absolute_error: 5.4737 - val_loss: 78.2141 - val_mean_absolute_error: 5.2664\n",
      "Epoch 37/50\n",
      "45/45 [==============================] - 1s 11ms/step - loss: 95.7900 - mean_absolute_error: 5.4224 - val_loss: 68.6313 - val_mean_absolute_error: 4.6680\n",
      "Epoch 38/50\n",
      "45/45 [==============================] - 0s 11ms/step - loss: 84.8300 - mean_absolute_error: 4.9451 - val_loss: 69.0069 - val_mean_absolute_error: 5.0291\n",
      "Epoch 39/50\n",
      "45/45 [==============================] - 1s 18ms/step - loss: 79.1025 - mean_absolute_error: 4.7375 - val_loss: 55.8874 - val_mean_absolute_error: 3.9746\n",
      "Epoch 40/50\n",
      "45/45 [==============================] - 1s 16ms/step - loss: 73.5489 - mean_absolute_error: 4.6576 - val_loss: 54.1457 - val_mean_absolute_error: 3.7675\n",
      "Epoch 41/50\n",
      "45/45 [==============================] - 0s 8ms/step - loss: 66.1009 - mean_absolute_error: 4.3989 - val_loss: 57.1837 - val_mean_absolute_error: 4.3764\n",
      "Epoch 42/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 58.7255 - mean_absolute_error: 3.8406 - val_loss: 46.5950 - val_mean_absolute_error: 3.5461\n",
      "Epoch 43/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 56.0743 - mean_absolute_error: 3.9895 - val_loss: 45.5377 - val_mean_absolute_error: 3.6420\n",
      "Epoch 44/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 57.9854 - mean_absolute_error: 4.2243 - val_loss: 48.9058 - val_mean_absolute_error: 3.9522\n",
      "Epoch 45/50\n",
      "45/45 [==============================] - 0s 5ms/step - loss: 50.8447 - mean_absolute_error: 3.6366 - val_loss: 43.3516 - val_mean_absolute_error: 3.6183\n",
      "Epoch 46/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 47.3967 - mean_absolute_error: 3.5179 - val_loss: 48.4901 - val_mean_absolute_error: 4.2452\n",
      "Epoch 47/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 44.0234 - mean_absolute_error: 3.4773 - val_loss: 48.8250 - val_mean_absolute_error: 4.1462\n",
      "Epoch 48/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 38.5991 - mean_absolute_error: 3.0879 - val_loss: 35.1132 - val_mean_absolute_error: 3.0639\n",
      "Epoch 49/50\n",
      "45/45 [==============================] - 0s 6ms/step - loss: 34.7008 - mean_absolute_error: 3.0121 - val_loss: 52.8159 - val_mean_absolute_error: 4.8251\n",
      "Epoch 50/50\n",
      "45/45 [==============================] - 0s 7ms/step - loss: 32.8030 - mean_absolute_error: 2.8786 - val_loss: 34.4224 - val_mean_absolute_error: 3.4093\n",
      "14/14 [==============================] - 0s 3ms/step - loss: 84762.8906 - mean_absolute_error: 259.8972\n",
      "Test Loss: 84762.890625, Test Mean Absolute Error: 259.8972473144531\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# get the best hyperparameters from the tuner\n",
    "best_hp = tuner.get_best_hyperparameters(1)[0]\n",
    "\n",
    "# build a new model using the best hyperparameters\n",
    "best_model = tuner.hypermodel.build(best_hp)\n",
    "\n",
    "# compile the best model\n",
    "best_model.compile(loss='mean_squared_error', optimizer='adam', metrics=['mean_absolute_error'])\n",
    "\n",
    "# train the best model with the training data\n",
    "best_history = best_model.fit(X_train_normalized, y_train, epochs=50, batch_size=32, validation_split=0.2)\n",
    "\n",
    "# evaluate the best model on your test data\n",
    "test_loss, test_mean_absolute_error = best_model.evaluate(X_test_normalized, y_test)\n",
    "print(f\"Test Loss: {test_loss}, Test Mean Absolute Error: {test_mean_absolute_error}\")\n",
    "\n",
    "# plot training & validation loss values\n",
    "plt.figure(figsize=(12, 6))\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.plot(best_history.history['loss'], label='Training Loss')\n",
    "plt.plot(best_history.history['val_loss'], label='Validation Loss')\n",
    "plt.title('Model Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend(['Train', 'Validation'], loc='upper right')\n",
    "\n",
    "# plot training & validation mean absolute error values (if using MAE as a metric)\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.plot(best_history.history['mean_absolute_error'], label='Training MAE')\n",
    "plt.plot(best_history.history['val_mean_absolute_error'], label='Validation MAE')\n",
    "plt.title('Mean Absolute Error')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('MAE')\n",
    "plt.legend(['Train', 'Validation'], loc='upper right')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "x8_rhH-8BSmX"
   },
   "source": [
    "A strong indication of a well-fitting model can be diagnosed from a plot in which both the training and validation losses steadily decrease and converge at a similar point.\n",
    "In such cases, the model demonstrates strong performance on both the training and validation datasets."
   ]
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}