course_completion_prediction_kaggle.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "course-completion-prediction-kaggle.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dZy8YK5PV-Kk"
      },
      "source": [
        "\n",
        "# Course Completion Prediction (On Kaggle)\n",
        "##CSE 416 Introduction to Machine Learning \n",
        "##Assignment 5 \n",
        "\n",
        "Sophia He \n",
        "\n",
        "July 28, 2020"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PFYlFCG8T1fg"
      },
      "source": [
        "This is a course-wide competition on Kaggle for CSE 416 at the University of Washington, via https://www.kaggle.com/c/cse-stat-416-sum-20/data. The objective is to investigate edX (anonymized) student data to try to predict whether a student received a certification (completed) a course. In this iPython Notebook, I explain my methods for data investigation, feature selection, and model training and evaluation."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "O1zCKmFiVf8T"
      },
      "source": [
        "import requests\n",
        "\n",
        "def save_file(url, file_name):\n",
        "    r = requests.get(url)\n",
        "    with open(file_name, 'wb') as f:\n",
        "        f.write(r.content)\n",
        "    \n",
        "save_file('https://homes.cs.washington.edu/~vinitra/cse416/a5/edx_train.csv', 'edx_train.csv') \n",
        "save_file('https://homes.cs.washington.edu/~vinitra/cse416/a5/edx_test.csv', 'edx_test.csv')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6RLtBP88V6kQ"
      },
      "source": [
        "import warnings\n",
        "\n",
        "import matplotlib.pyplot as plt # for graphing\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import seaborn as sns\n",
        "from mpl_toolkits.mplot3d import Axes3D\n",
        "\n",
        "%matplotlib inline\n",
        "warnings.filterwarnings('ignore')\n",
        "sns.set()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5imyOgv7Z6GB"
      },
      "source": [
        "# Investigate Data\n",
        "\n",
        "I started out by importing the test and train data and looking at the train data. I verified that the training data has an extra 'certified' feature that the test data does not have, which will be the target."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8I_EjG5MZ5Xc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 72
        },
        "outputId": "8dbd7968-2870-4053-ab1d-049e8864ca37"
      },
      "source": [
        "edx_data = pd.read_csv('edx_train.csv')\n",
        "edx_test = pd.read_csv('edx_test.csv')    # csv file with only features\n",
        "print(list(edx_data.columns))\n",
        "\n",
        "print(list(edx_test.columns))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['course_id', 'userid_DI', 'registered', 'viewed', 'explored', 'certified', 'final_cc_cname_DI', 'LoE_DI', 'YoB', 'gender', 'start_time_DI', 'last_event_DI', 'nevents', 'ndays_act', 'nplay_video', 'nchapters', 'nforum_posts']\n",
            "['course_id', 'userid_DI', 'registered', 'viewed', 'explored', 'final_cc_cname_DI', 'LoE_DI', 'YoB', 'gender', 'start_time_DI', 'last_event_DI', 'nevents', 'ndays_act', 'nplay_video', 'nchapters', 'nforum_posts']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eIi2hDvKaSTX",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 217
        },
        "outputId": "4b076710-a7f3-45ae-95e8-d5ea7aa9d452"
      },
      "source": [
        "edx_data.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>course_id</th>\n",
              "      <th>userid_DI</th>\n",
              "      <th>registered</th>\n",
              "      <th>viewed</th>\n",
              "      <th>explored</th>\n",
              "      <th>certified</th>\n",
              "      <th>final_cc_cname_DI</th>\n",
              "      <th>LoE_DI</th>\n",
              "      <th>YoB</th>\n",
              "      <th>gender</th>\n",
              "      <th>start_time_DI</th>\n",
              "      <th>last_event_DI</th>\n",
              "      <th>nevents</th>\n",
              "      <th>ndays_act</th>\n",
              "      <th>nplay_video</th>\n",
              "      <th>nchapters</th>\n",
              "      <th>nforum_posts</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>HarvardX/CS50x/2012</td>\n",
              "      <td>MHxPC130288921</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>Unknown/Other</td>\n",
              "      <td>Bachelor's</td>\n",
              "      <td>1988.0</td>\n",
              "      <td>m</td>\n",
              "      <td>8/30/12</td>\n",
              "      <td>unk</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>HarvardX/PH207x/2012_Fall</td>\n",
              "      <td>MHxPC130043703</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>India</td>\n",
              "      <td>Doctorate</td>\n",
              "      <td>1970.0</td>\n",
              "      <td>m</td>\n",
              "      <td>9/4/12</td>\n",
              "      <td>1/16/13</td>\n",
              "      <td>2550.0</td>\n",
              "      <td>21.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>14.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>HarvardX/CS50x/2012</td>\n",
              "      <td>MHxPC130082773</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>United States</td>\n",
              "      <td>Master's</td>\n",
              "      <td>1983.0</td>\n",
              "      <td>m</td>\n",
              "      <td>2/12/13</td>\n",
              "      <td>2/12/13</td>\n",
              "      <td>10.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>HarvardX/PH207x/2012_Fall</td>\n",
              "      <td>MHxPC130186614</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>Other Africa</td>\n",
              "      <td>Bachelor's</td>\n",
              "      <td>1984.0</td>\n",
              "      <td>m</td>\n",
              "      <td>10/16/12</td>\n",
              "      <td>1/15/13</td>\n",
              "      <td>6015.0</td>\n",
              "      <td>50.0</td>\n",
              "      <td>616.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>HarvardX/CS50x/2012</td>\n",
              "      <td>MHxPC130113075</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>United States</td>\n",
              "      <td>Secondary</td>\n",
              "      <td>1997.0</td>\n",
              "      <td>m</td>\n",
              "      <td>4/11/13</td>\n",
              "      <td>4/12/13</td>\n",
              "      <td>6.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                   course_id       userid_DI  ...  nchapters  nforum_posts\n",
              "0        HarvardX/CS50x/2012  MHxPC130288921  ...        0.0             0\n",
              "1  HarvardX/PH207x/2012_Fall  MHxPC130043703  ...       14.0             0\n",
              "2        HarvardX/CS50x/2012  MHxPC130082773  ...        1.0             0\n",
              "3  HarvardX/PH207x/2012_Fall  MHxPC130186614  ...       15.0             0\n",
              "4        HarvardX/CS50x/2012  MHxPC130113075  ...        1.0             0\n",
              "\n",
              "[5 rows x 17 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1244
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qMxCEcPdeopO"
      },
      "source": [
        "I then looked at distribution plots for a couple of features in the training data, to better visualize the data categories."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Hhb2ZgHJaoUd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 302
        },
        "outputId": "8d0181d1-630f-4d06-ce20-6af779aacd74"
      },
      "source": [
        "sns.countplot(x='gender', data=edx_data, order=sorted(edx_data['gender'].unique()))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fae4c5b9e48>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1245
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZMAAAEMCAYAAAABLFv3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAVn0lEQVR4nO3de3BU9d3H8c/uwgYQwpKQhE1wKmjLZIyYQqy9gAoMBjsB4oUmk6coIq2KKWKHSwo26YRLm0vreGGwnaDMVNq0Y0s0wRKKsVN1qBbaaEMstmm0pdkmsCECGhLInucPhx0zIiz7S85mN+/XjGN2f3v5hp3wzjmHPeuwLMsSAAAGnJEeAAAQ/YgJAMAYMQEAGCMmAABjxAQAYIyYAACMERMAgLERkR4gkk6c+FCBAG+zAYBQOJ0OTZhwxQXXhnVMAgGLmADAAGA3FwDAGDEBABgjJgAAY8QEAGCMmAAAjBETAIAxYgIAMDas32eC2DdhvFsj3HGRHiPmnevt0YkPeiM9BiKImCCmjXDH6VD5ikiPEfNmrquSREyGM3ZzAQCM2bZlMnfuXLndbsXFfbzLYc2aNZo9e7YaGxtVXFysnp4epaWlqaKiQomJiZIU9hoAwF62bpk88cQTeuGFF/TCCy9o9uzZCgQCWrt2rYqLi1VfX6+srCxVVlZKUthrAAD7RXQ3V1NTk+Li4pSVlSVJys/P1969e43WAAD2s/UA/Jo1a2RZlmbOnKnvfve78vl8Sk1NDa4nJCQoEAioq6sr7DWPxxPyPImJYwfmGwOgpKRxkR4BEWRbTHbt2iWv16ve3l5t2bJFpaWlmj9/vl1Pf0F+/2lOQR/j+AvOPseOnYr0CBhkTqfjM38Jt203l9frlSS53W4VFBToL3/5i7xer9ra2oK36ezslNPplMfjCXsNAGA/W2Ly0Ucf6dSpj39rsSxLL730ktLT05WRkaEzZ87o4MGDkqTq6motWLBAksJeAwDYz5bdXH6/X9/5znfU19enQCCgq6++WiUlJXI6nSovL1dJSUm/f+IrKew1AID9HJZlDduDBhwziX1JSeN4B7wNZq6r4pjJMDAkjpkAAGIXMQEAGCMmAABjxAQAYIyYAACMERMAgDFiAgAwRkwAAMaICQDAGDEBABgjJgAAY8QEAGCMmAAAjBETAIAxYgIAMEZMAADGiAkAwBgxAQAYIyYAAGPEBABgjJgAAIwREwCAMWICADBGTAAAxogJAMAYMQEAGCMmAABjxAQAYIyYAACMERMAgDFiAgAwZntMnnrqKU2bNk3vvvuuJKmxsVGLFi1Sdna2li9fLr/fH7xtuGsAAHvZGpPDhw+rsbFRaWlpkqRAIKC1a9equLhY9fX1ysrKUmVlpdEaAMB+tsWkt7dXpaWl+sEPfhC8rqmpSXFxccrKypIk5efna+/evUZrAAD7jbDriR5//HEtWrRIkydPDl7n8/mUmpoavJyQkKBAIKCurq6w1zweT8gzJSaONfyuAJyXlDQu0iMggmyJyV//+lc1NTVpzZo1djxdyPz+0woErEiPgUHEX3D2OXbsVKRHwCBzOh2f+Uu4LTH585//rJaWFs2bN0+S9L///U/33Xefli5dqra2tuDtOjs75XQ65fF45PV6w1oDANjPlmMm3/72t/Xaa6+poaFBDQ0NmjRpknbs2KEVK1bozJkzOnjwoCSpurpaCxYskCRlZGSEtQYAsJ9tx0wuxOl0qry8XCUlJerp6VFaWpoqKiqM1gAA9nNYljVsDxpwzCT2JSWN06HyFZEeI+bNXFfFMZNh4GLHTHgHPADAGDEBABgjJgAAY8QEAGCMmAAAjBETAIAxYgIAMEZMAADGiAkAwBgxAQAYIyYAAGPEBABgjJgAAIwREwCAMWICADBGTAAAxogJAMAYMQEAGCMmAABjxAQAYIyYAACMERMAgDFiAgAwRkwAAMaICQDAGDEBABgjJgAAY8QEAGCMmAAAjBETAIAxYgIAMDbCridauXKljh49KqfTqTFjxuj73/++0tPT1draqqKiInV1dcnj8aisrExXXXWVJIW9BgCwV8hbJjt27Ljg9c8++2xI9y8rK9OLL76ompoaLV++XBs2bJAklZSUqKCgQPX19SooKFBxcXHwPuGuAQDsFXJMtm3bdsHrt2/fHtL9x40bF/z69OnTcjgc8vv9am5uVk5OjiQpJydHzc3N6uzsDHsNAGC/S+7mOnDggCQpEAjoT3/6kyzLCq4dPXpUV1xxRchPtnHjRr3++uuyLEtVVVXy+XxKSUmRy+WSJLlcLiUnJ8vn88myrLDWEhISQv/uAQAD4pIx2bhxoySpp6cnuGtKkhwOh5KSkvToo4+G/GRbtmyRJNXU1Ki8vFwPP/zw5c47oBITx0b0+YFYkpQ07tI3Qsy6ZEwaGhokSevWrVN5efmAPGlubq6Ki4s1adIktbe3q6+vTy6XS319fero6JDX65VlWWGtXQ6//7QCAevSN0TU4i84+xw7dirSI2CQOZ2Oz/wlPORjJp8MSSAQ6PffpXz44Yfy+XzByw0NDRo/frwSExOVnp6uuro6SVJdXZ3S09OVkJAQ9hoAwH4O65MHQS7i8OHDKi0t1ZEjR9TT0yNJsixLDodD77zzzkXve/z4ca1cuVLd3d1yOp0aP3681q9fr2uvvVYtLS0qKirSyZMnFR8fr7KyMk2dOlWSwl4LFVsmsS8paZwOla+I9Bgxb+a6KrZMhoGLbZmEHJOFCxdqzpw5Wrx4sUaNGtVvLS0tzXzKCCAmsY+Y2IOYDA8Xi0nIb1r873//q0ceeUQOh2PABgMAxIaQj5nMnz9fr7322mDOAgCIUiFvmfT09KiwsFAzZ87UxIkT+60N1L/yGqrGxY/SqLiRkR4jpp3pOatTJ89EegwAYQo5Jtdcc42uueaawZxlyBoVN1IF63ZFeoyY9ovy/9MpERMgWoUck8LCwsGcAwAQxUKOyfnTqlzIV77ylQEZBgAQnUKOyfnTqpx34sQJnT17VikpKXr55ZcHfDAAQPQIOSbnT6tyXl9fn7Zv335ZJ3oEAMSmsD9p0eVy6YEHHlBVVdVAzgMAiEJGH9v7+uuv8yZGAEDou7luvvnmfuHo7u5Wb2+vSkpKBmUwAED0CDkmFRUV/S6PHj1aU6ZM0dixfCYIAAx3IcfkS1/6kqSPTz9//PhxTZw4UU6n0V4yAECMCLkGp0+f1rp16zR9+nTddNNNmj59utavX69TpzhTKAAMdyHHZPPmzeru7lZtba3efvtt1dbWqru7W5s3bx7M+QAAUSDk3Vyvvvqq9u/fr9GjR0uSpkyZoh/+8IeaP3/+oA0HAIgOIW+ZxMXFqbOzs991J06ckNvtHvChAADRJeQtk7vuukvLly/XsmXLlJqaqra2Nu3cuVNLliwZzPkAAFEg5Jg8+OCDSklJUW1trTo6OpScnKwVK1YQEwBA6Lu5tmzZoilTpmjnzp166aWXtHPnTl199dXasmXLYM4HAIgCIcekrq5OGRkZ/a7LyMhQXV3dgA8FAIguIcfE4XAoEAj0u66vr+9T1wEAhp+QY5KVlaXHH388GI9AIKAnn3xSWVlZgzYcACA6XNaHY91///2aNWuWUlNT5fP5lJSUpKeffnow5wMARIGQYzJp0iTt3r1bb7/9tnw+n7xer6ZPn875uQAAocdEkpxOpzIzM5WZmTlY8wAAohCbFQAAY8QEAGCMmAAAjBETAIAxYgIAMGZLTE6cOKFvfetbys7O1sKFC1VYWBg8nX1jY6MWLVqk7OxsLV++XH6/P3i/cNcAAPayJSYOh0MrVqxQfX29amtrdeWVV6qyslKBQEBr165VcXGx6uvrlZWVpcrKSkkKew0AYD9bYuLxeHTjjTcGL2dmZqqtrU1NTU2Ki4sLnpIlPz9fe/fulaSw1wAA9rP9mEkgENAvf/lLzZ07Vz6fT6mpqcG1hIQEBQIBdXV1hb0GALDfZb0DfiBs2rRJY8aM0Te/+U39/ve/t/vp+0lMHBvR50d/SUnjIj0CDPD6DW+2xqSsrEzvv/++nn76aTmdTnm9XrW1tQXXOzs75XQ65fF4wl67HH7/aQUC1iVvxw+JPY4dOzXgj8lrZ5/BeP0wtDidjs/8Jdy23Vw/+clP1NTUpG3btsntdkv6+MO1zpw5o4MHD0qSqqurtWDBAqM1AID9bNky+cc//qGf/vSnuuqqq5Sfny9Jmjx5srZt26by8nKVlJSop6dHaWlpqqiokPTxSSXDWQMA2M+WmHz+85/XkSNHLrg2Y8YM1dbWDugaAMBevAMeAGCMmAAAjBETAIAxYgIAMEZMAADGiAkAwBgxAQAYIyYAAGPEBABgjJgAAIwREwCAMWICADBGTAAAxogJAMAYMQEAGCMmAABjxAQAYIyYAACMERMAgDFiAgAwRkwAAMaICQDAGDEBABgjJgAAY8QEAGCMmAAAjBETAIAxYgIAMEZMAADGiAkAwBgxAQAYsyUmZWVlmjt3rqZNm6Z33303eH1ra6vy8vKUnZ2tvLw8vffee8ZrAAD72RKTefPmadeuXUpLS+t3fUlJiQoKClRfX6+CggIVFxcbrwEA7GdLTLKysuT1evtd5/f71dzcrJycHElSTk6Ompub1dnZGfYaACAyRkTqiX0+n1JSUuRyuSRJLpdLycnJ8vl8siwrrLWEhIRIfTsAMKxFLCZDQWLi2EiPgE9IShoX6RFggNdveItYTLxer9rb29XX1yeXy6W+vj51dHTI6/XKsqyw1i6X339agYB1ydvxQ2KPY8dODfhj8trZZzBePwwtTqfjM38Jj9g/DU5MTFR6errq6uokSXV1dUpPT1dCQkLYawCAyHBYlnXpX80Nbd68Wfv27dPx48c1YcIEeTwe7dmzRy0tLSoqKtLJkycVHx+vsrIyTZ06VZLCXrscl7NlUrBu12U/PkL3i/L/G7Qtk0PlKwb8cdHfzHVVbJkMAxfbMrElJkMVMRk6iEl0IybDw5DczQUAiB3EBABgjJgAAIwREwCAsWH9pkUAQ1v8+DjFud2RHiOm9fT26uQHPcaPQ0wADFlxbreWPftwpMeIaTvvfVySeUzYzQUAMEZMAADGiAkAwBgxAQAYIyYAAGPEBABgjJgAAIwREwCAMWICADBGTAAAxogJAMAYMQEAGCMmAABjxAQAYIyYAACMERMAgDFiAgAwRkwAAMaICQDAGDEBABgjJgAAY8QEAGCMmAAAjBETAIAxYgIAMEZMAADGojomra2tysvLU3Z2tvLy8vTee+9FeiQAGJaiOiYlJSUqKChQfX29CgoKVFxcHOmRAGBYGhHpAcLl9/vV3NysZ599VpKUk5OjTZs2qbOzUwkJCSE9htPpCPn5Jk64Iqw5EbrLeT0uhzs+cVAeF/0N1us3cWxoP88IX6iv3cVu57AsyxqogezU1NSk9evXa8+ePcHrvv71r6uiokLXXnttBCcDgOEnqndzAQCGhqiNidfrVXt7u/r6+iRJfX196ujokNfrjfBkADD8RG1MEhMTlZ6errq6OklSXV2d0tPTQz5eAgAYOFF7zESSWlpaVFRUpJMnTyo+Pl5lZWWaOnVqpMcCgGEnqmMCABgaonY3FwBg6CAmAABjxAQAYIyYAACMEZMYtH//ft12223Kzc3Vv/71r0iPAwxbR48e1Y033hjpMWwRtefmwmerrq7WqlWrdNttt0V6FADDBFsmMWbr1q06dOiQKisrtXTp0kiPgxBMmzZN27dv15133ql58+bpwIED+vGPf6zc3Fzl5OSopaUl0iNCn97KOH/5/P8fe+wx5ebmKjs7WwcPHvzU/Xt7e7V69Wr96Ec/Uiy+I4OYxJgNGzYoIyNDjz76qH7+859HehyEKD4+Xr/5zW+0Zs0arVy5UjNmzFBNTY0WL16s7du3R3o8XEJXV5cyMzNVU1Ojhx56SJWVlZ9av++++zRjxgwVFRXJ4RicMyxHEjEBhoDzuyTPn/F6zpw5kqSMjAz9+9//jthcCM2YMWOCr1lmZqb+85//BNd6e3tVUFCggoIC3X333ZEacdARE2AIiIuLkyQ5nU653e7g9U6nU+fOnYvUWPiEESNG9Ns91dPTE/z6Yq/ZyJEjdf3116uhoSF4YtpYREwAIAQTJ07U2bNn9f7770tS8CSzl+JwOLR161aNHTtWjzzyiM6ePTuYY0YMMQGAEIwYMUIbN27Uvffeq7vuuksulyvk+zocDpWUlCgtLU0PPfRQv62aWMGJHgEAxtgyAQAYIyYAAGPEBABgjJgAAIwREwCAMWICRIlp06YF3+MADDXEBABgjJgAMY7TscAOxAQwdPjwYeXm5uqLX/yiVq1apdWrV+uxxx6TJL3yyitavHixsrKylJ+fr7///e/B+82dO1c7duzQwoULNXPmTK1evbrfO6Orqqo0a9YszZo1S88//3y/5+zt7VVZWZluueUWffWrX1VxcbHOnDkjSXrjjTd000036Wc/+5m+9rWv6Xvf+54NfwoY7ogJYKC3t1eFhYW6/fbb9eabbyonJ0f79++XJDU3N2vDhg0qLS3VG2+8oby8PK1cuVK9vb3B+//ud79TVVWVXn75ZR05ckS//e1vJUl//OMf9cwzz+iZZ57Rvn37dODAgX7PW1lZqdbWVtXU1Gjfvn3q6OjQtm3bguvHjx/XBx98oFdeeUWbNm2y4U8Cwx0xAQy89dZbOnfunO6++26NHDlSt956q6677jpJ0q9+9Svl5eXp+uuvl8vl0u23366RI0eqsbExeP+lS5cqJSVFHo9Hc+bM0TvvvCPp48jccccd+sIXvqAxY8aosLAweB/LsvTrX/9aGzZskMfj0dixY3X//fdrz549wds4nU6tWrVKbrdbo0aNsulPA8MZH9sLGOjo6FBKSkq/Dzvyer2SpLa2NtXU1Oi5554Lrp09e1YdHR3By0lJScGvR48eHVzr6OhQRkZGcC0tLS34dWdnp7q7u3XHHXcEr7MsS4FAIHh5woQJwdPaA3YgJoCBpKQktbe3y7KsYFB8Pp+uvPJKeb1ePfDAA3rwwQcv+3GTk5Pl8/mCl9va2oJfT5gwQaNGjdKePXuUkpJywfvH4if5YWhjNxdgIDMzUy6XS88995zOnTun/fv3629/+5skacmSJaqurtZbb70ly7L00Ucf6Q9/+INOnz59ycddsGCBdu/erX/+85/q7u7WU089FVxzOp1asmSJtm7dKr/fL0lqb2/Xq6++OjjfJBACYgIYcLvdevLJJ/X888/rhhtu0IsvvqhbbrlFbrdb1113nTZt2qTS0lLdcMMNuvXWW4MH2C/l5ptv1j333KN77rlH8+fP15e//OV+62vXrtXnPvc5feMb39CMGTO0bNkytba2Dsa3CISEzzMBBtiSJUuUn5+vO++8M9KjALZhywQw9Oabb+rYsWM6d+6cdu/erSNHjmj27NmRHguwFQfgAUOtra1avXq1uru7NXnyZD3xxBNKTk6O9FiArdjNBQAwxm4uAIAxYgIAMEZMAADGiAkAwBgxAQAYIyYAAGP/D9cOOFpUM2CVAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FsiIjPQ3eXK4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 302
        },
        "outputId": "547d2b81-dcb8-4b81-f5a2-845f5d2f9320"
      },
      "source": [
        "sns.countplot(x='viewed', data=edx_data, order=sorted(edx_data['viewed'].unique()))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fae4c0bb710>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1246
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZPGk3ra_fhoj"
      },
      "source": [
        "In the training data, I looked at the split between certified and not certified, in the column of the target feature."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_IDQMmHUdJVW",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "d3d5d913-8cdd-438d-875f-ac9ca53f1869"
      },
      "source": [
        "only_certified = edx_data[edx_data['certified'] == 1]\n",
        "only_not_certified = edx_data[edx_data['certified'] == 0]\n",
        "\n",
        "print(f'Number certified: {len(only_certified)} ({len(only_certified) * 100.0 / len(edx_data):.2f}%)')\n",
        "print(f'Number not certified: {len(only_not_certified)} ({len(only_not_certified) * 100.0 / len(edx_data):.2f}%)')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Number certified: 3826 (43.69%)\n",
            "Number not certified: 4932 (56.31%)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mabrKq8hf1TO"
      },
      "source": [
        "# Feature Selection"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_kyMyAkPfyOJ"
      },
      "source": [
        "To start off with feature selection, I listed all the features in the training data."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bX3mVGCEpNsz",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 318
        },
        "outputId": "fbcbc74c-a8da-45cb-ff90-ee9b1053157d"
      },
      "source": [
        "list(edx_data.columns)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['course_id',\n",
              " 'userid_DI',\n",
              " 'registered',\n",
              " 'viewed',\n",
              " 'explored',\n",
              " 'certified',\n",
              " 'final_cc_cname_DI',\n",
              " 'LoE_DI',\n",
              " 'YoB',\n",
              " 'gender',\n",
              " 'start_time_DI',\n",
              " 'last_event_DI',\n",
              " 'nevents',\n",
              " 'ndays_act',\n",
              " 'nplay_video',\n",
              " 'nchapters',\n",
              " 'nforum_posts']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1248
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZjHi6yopgG8a"
      },
      "source": [
        "Following, I defined a features array (shown below) and commented out the features to exclude. Initially, the only features excluded were 'userid_DI' (as suggested on the Kaggle page), as well as 'start_time_DI' because of the high variation from date to date (too many categories) and 'last_event_DI' for the same reason, as well as because nevents already gave a metric related to interaction with the course. \n",
        "\n",
        "After creating my first model (a decision tree) I came back and commented out two other features that were negatively impacting accuracy, 'final_cc_cname_DI' and 'LoE_DI'.\n",
        "\n",
        "I chose not to reformat the data and keep the 'unk' and -1 representing unknowns as is because for many features there was no way to average them or find an appropriate way to recategorize them (such as with gender). I did not want to place any assumptions that would mis-categorize all unknowns for a feature and skew the data.\n",
        "\n",
        "I then tried three more models: Random Forest, AdaBoost, and K Nearest Neighbor. I compared the resulting validation accuracies from these three models with the decision tree with the max depth with the best accuracy, and I found that Random Forest resulted in the best accuracy overall. I then changed the features once more to attain the best possible Random Forest accuracy, resulting in the features you see below. The excluded features are commented out."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "erhK2AD1ojz1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 217
        },
        "outputId": "719bd8f6-8cb5-49a6-d1d7-2f9b24a7c1db"
      },
      "source": [
        "# Feature Set\n",
        "features = [  \n",
        "    'course_id',                # institution, course name, semester\n",
        "    #'userid_DI',               # id number\n",
        "    'registered',               # 0/1 registered for course = 1\n",
        "    'viewed',                   # 0/1 anyone  who accessed  the ‘Courseware’  tab \n",
        "    'explored',                 # 0/1;  anyone  who accessed  at  least half  of  the chapters\n",
        "    #'final_cc_cname_DI',       # country or continent/region\n",
        "    'LoE_DI',                  # user provided highest level of education\n",
        "    'YoB',                      # user provided year of birth\n",
        "    'gender',                   # user provided gender m, f, or o\n",
        "    #'start_time_DI',           # date of course registration\n",
        "    #'last_event_DI',           # date of last interaction with course\n",
        "    'nevents',                  # number of interactions with the course\n",
        "    'ndays_act',                # number of unique days student interacted with course\n",
        "    'nplay_video',              # number of play video  events within the course\n",
        "    'nchapters',                # number of chapters with which the student interacted\n",
        "    'nforum_posts',             # number of posts to the Discussion Forum. \n",
        "]\n",
        "\n",
        "\n",
        "target = 'certified'                   # prediction target (y) (1 means certified, 0 is not certified)\n",
        "\n",
        "# Extract the feature columns and target column\n",
        "edx_data = edx_data[features + [target]]\n",
        "edx_data.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>course_id</th>\n",
              "      <th>registered</th>\n",
              "      <th>viewed</th>\n",
              "      <th>explored</th>\n",
              "      <th>LoE_DI</th>\n",
              "      <th>YoB</th>\n",
              "      <th>gender</th>\n",
              "      <th>nevents</th>\n",
              "      <th>ndays_act</th>\n",
              "      <th>nplay_video</th>\n",
              "      <th>nchapters</th>\n",
              "      <th>nforum_posts</th>\n",
              "      <th>certified</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>HarvardX/CS50x/2012</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>Bachelor's</td>\n",
              "      <td>1988.0</td>\n",
              "      <td>m</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>HarvardX/PH207x/2012_Fall</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>Doctorate</td>\n",
              "      <td>1970.0</td>\n",
              "      <td>m</td>\n",
              "      <td>2550.0</td>\n",
              "      <td>21.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>14.0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>HarvardX/CS50x/2012</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>Master's</td>\n",
              "      <td>1983.0</td>\n",
              "      <td>m</td>\n",
              "      <td>10.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>HarvardX/PH207x/2012_Fall</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>Bachelor's</td>\n",
              "      <td>1984.0</td>\n",
              "      <td>m</td>\n",
              "      <td>6015.0</td>\n",
              "      <td>50.0</td>\n",
              "      <td>616.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>HarvardX/CS50x/2012</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>Secondary</td>\n",
              "      <td>1997.0</td>\n",
              "      <td>m</td>\n",
              "      <td>6.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                   course_id  registered  ...  nforum_posts  certified\n",
              "0        HarvardX/CS50x/2012           1  ...             0          0\n",
              "1  HarvardX/PH207x/2012_Fall           1  ...             0          1\n",
              "2        HarvardX/CS50x/2012           1  ...             0          0\n",
              "3  HarvardX/PH207x/2012_Fall           1  ...             0          1\n",
              "4        HarvardX/CS50x/2012           1  ...             0          0\n",
              "\n",
              "[5 rows x 13 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1249
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CUcKbGFGmOev"
      },
      "source": [
        "Because many categories had non-numerical data, for preprocessing I used one-hot encoding via the get_dummies() method; though it introduced sparsity into the dataset, it resulted in numerical values for the non-numerical categories."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7MhdACl8qOxS",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 425
        },
        "outputId": "38e4eca6-4ca0-4636-f5eb-44f5b495c8fe"
      },
      "source": [
        "edx_data = pd.get_dummies(edx_data)\n",
        "features = list(edx_data.columns)\n",
        "features.remove('certified')\n",
        "features"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['registered',\n",
              " 'viewed',\n",
              " 'explored',\n",
              " 'YoB',\n",
              " 'nevents',\n",
              " 'ndays_act',\n",
              " 'nplay_video',\n",
              " 'nchapters',\n",
              " 'nforum_posts',\n",
              " 'course_id_HarvardX/CB22x/2013_Spring',\n",
              " 'course_id_HarvardX/CS50x/2012',\n",
              " 'course_id_HarvardX/ER22x/2013_Spring',\n",
              " 'course_id_HarvardX/PH207x/2012_Fall',\n",
              " 'course_id_HarvardX/PH278x/2013_Spring',\n",
              " \"LoE_DI_Bachelor's\",\n",
              " 'LoE_DI_Doctorate',\n",
              " 'LoE_DI_Less than Secondary',\n",
              " \"LoE_DI_Master's\",\n",
              " 'LoE_DI_Secondary',\n",
              " 'LoE_DI_unk',\n",
              " 'gender_f',\n",
              " 'gender_m',\n",
              " 'gender_unk']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1250
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mt7m6ijtryTF",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 251
        },
        "outputId": "1addd058-b42c-412e-e118-6d09b008b8d8"
      },
      "source": [
        "edx_data.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>registered</th>\n",
              "      <th>viewed</th>\n",
              "      <th>explored</th>\n",
              "      <th>YoB</th>\n",
              "      <th>nevents</th>\n",
              "      <th>ndays_act</th>\n",
              "      <th>nplay_video</th>\n",
              "      <th>nchapters</th>\n",
              "      <th>nforum_posts</th>\n",
              "      <th>certified</th>\n",
              "      <th>course_id_HarvardX/CB22x/2013_Spring</th>\n",
              "      <th>course_id_HarvardX/CS50x/2012</th>\n",
              "      <th>course_id_HarvardX/ER22x/2013_Spring</th>\n",
              "      <th>course_id_HarvardX/PH207x/2012_Fall</th>\n",
              "      <th>course_id_HarvardX/PH278x/2013_Spring</th>\n",
              "      <th>LoE_DI_Bachelor's</th>\n",
              "      <th>LoE_DI_Doctorate</th>\n",
              "      <th>LoE_DI_Less than Secondary</th>\n",
              "      <th>LoE_DI_Master's</th>\n",
              "      <th>LoE_DI_Secondary</th>\n",
              "      <th>LoE_DI_unk</th>\n",
              "      <th>gender_f</th>\n",
              "      <th>gender_m</th>\n",
              "      <th>gender_unk</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1988.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1970.0</td>\n",
              "      <td>2550.0</td>\n",
              "      <td>21.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>14.0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1983.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1984.0</td>\n",
              "      <td>6015.0</td>\n",
              "      <td>50.0</td>\n",
              "      <td>616.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1997.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   registered  viewed  explored  ...  gender_f  gender_m  gender_unk\n",
              "0           1       0         0  ...         0         1           0\n",
              "1           1       1         1  ...         0         1           0\n",
              "2           1       1         0  ...         0         1           0\n",
              "3           1       1         1  ...         0         1           0\n",
              "4           1       1         0  ...         0         1           0\n",
              "\n",
              "[5 rows x 24 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1251
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e9s8jklPf8b9"
      },
      "source": [
        "# Model Training and Evaluation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ec7xSEx7pK5h"
      },
      "source": [
        "To train the models, I split the train data using train_test_split, to get both a train and validation dataset."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0_iLW3FOgEDW"
      },
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "train_data, validation_data = train_test_split(edx_data, test_size=0.2, random_state=1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lIU8ioMRpaW_"
      },
      "source": [
        "## Decision Tree Classifier\n",
        "\n",
        "I started out with the decision tree classifier, testing with a small tree of max depth 2, mid-sized tree with max depth 5, and big tree with max depth 10. Out of these three models, the mid-sized tree with max depth 6 resulted in the best validation data accuracy. As mentioned above, I then reselected and commented out two additional features 'final_cc_cname_DI' and 'LoE_DI', which increased the validation accuracy for the mid-sized tree."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZQ3Rn5mJsBMI",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 123
        },
        "outputId": "4f5ec29a-78d6-4fcb-b445-d942c6eaa1aa"
      },
      "source": [
        "from sklearn.tree import DecisionTreeClassifier\n",
        "# mid sized tree\n",
        "decision_tree_model = DecisionTreeClassifier(max_depth=5, random_state=1)\n",
        "decision_tree_model.fit(train_data[features], train_data[target])\n",
        "\n",
        "# small tree\n",
        "from sklearn import tree\n",
        "small_tree_model = DecisionTreeClassifier(max_depth=2, random_state=1)\n",
        "small_tree_model.fit(train_data[features], train_data[target])\n",
        "\n",
        "# big tree\n",
        "big_tree_model = DecisionTreeClassifier(max_depth=10, random_state=1)\n",
        "big_tree_model.fit(train_data[features], train_data[target])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "DecisionTreeClassifier(ccp_alpha=0.0, class_weight=None, criterion='gini',\n",
              "                       max_depth=10, max_features=None, max_leaf_nodes=None,\n",
              "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                       min_samples_leaf=1, min_samples_split=2,\n",
              "                       min_weight_fraction_leaf=0.0, presort='deprecated',\n",
              "                       random_state=1, splitter='best')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1253
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kBYz2o5ssdMJ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "4a6be04e-22e8-4388-fedf-a35e54419b9d"
      },
      "source": [
        "from sklearn.metrics import accuracy_score\n",
        "\n",
        "# small tree accuracy comparison\n",
        "small_tree_val_prediction = small_tree_model.predict(validation_data[features])\n",
        "small_tree_train_prediction = small_tree_model.predict(train_data[features])\n",
        "small_tree_val_acc = accuracy_score(small_tree_val_prediction, validation_data[target])\n",
        "small_tree_train_acc = accuracy_score(small_tree_train_prediction, train_data[target])\n",
        "\n",
        "print(\"small tree validation accuracy:\", small_tree_val_acc)\n",
        "print(\"small tree training accuracy:\", small_tree_train_acc)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "small tree validation accuracy: 0.9503424657534246\n",
            "small tree training accuracy: 0.9511846988295747\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iSx6TQXdsgq4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "2754461e-6593-4df0-ffd2-f766b26d97af"
      },
      "source": [
        "# mid sized tree accuracy comparison\n",
        "decision_tree_val_prediction = decision_tree_model.predict(validation_data[features])\n",
        "decision_tree_train_prediction = decision_tree_model.predict(train_data[features])\n",
        "decision_tree_val_acc = accuracy_score(decision_tree_val_prediction, validation_data[target])\n",
        "decision_tree_train_acc = accuracy_score(decision_tree_train_prediction, train_data[target])\n",
        "\n",
        "print(\"decision tree validation accuracy:\", decision_tree_val_acc)\n",
        "print(\"decision tree training accuracy:\", decision_tree_train_acc)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "decision tree validation accuracy: 0.9714611872146118\n",
            "decision tree training accuracy: 0.9737367970311162\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Wi6pl5SqsouO",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "dbfdad5f-281b-45f0-bb2a-c5f3546216b4"
      },
      "source": [
        "# big tree accuracy comparison\n",
        "big_tree_val_prediction = big_tree_model.predict(validation_data[features])\n",
        "big_tree_train_prediction = big_tree_model.predict(train_data[features])\n",
        "big_tree_val_acc = accuracy_score(big_tree_val_prediction, validation_data[target])\n",
        "big_tree_train_acc = accuracy_score(big_tree_train_prediction, train_data[target])\n",
        "\n",
        "print(\"big tree validation accuracy:\", big_tree_val_acc)\n",
        "print(\"big tree training accuracy:\", big_tree_train_acc)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "big tree validation accuracy: 0.9731735159817352\n",
            "big tree training accuracy: 0.9915786468741079\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cLn6L2_-rWR4"
      },
      "source": [
        "## Exporting Test Prediction for Decision Tree Classifier"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "35wcV1xIrsSc"
      },
      "source": [
        "I processed the test data in the same way as the training/validation data to prepare for getting model predictions."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TWgJTO28A74M",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 251
        },
        "outputId": "4b24e37b-5208-4f65-a21f-175e648ce283"
      },
      "source": [
        "target = 'certified'                   # prediction target (y) (+1 means certified, -1 is not certified)\n",
        "\n",
        "# Extract the feature columns and target column\n",
        "edx_id = edx_test[['userid_DI']]\n",
        "edx_test = pd.get_dummies(edx_test)\n",
        "edx_test = edx_test[features]\n",
        "edx_test.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>registered</th>\n",
              "      <th>viewed</th>\n",
              "      <th>explored</th>\n",
              "      <th>YoB</th>\n",
              "      <th>nevents</th>\n",
              "      <th>ndays_act</th>\n",
              "      <th>nplay_video</th>\n",
              "      <th>nchapters</th>\n",
              "      <th>nforum_posts</th>\n",
              "      <th>course_id_HarvardX/CB22x/2013_Spring</th>\n",
              "      <th>course_id_HarvardX/CS50x/2012</th>\n",
              "      <th>course_id_HarvardX/ER22x/2013_Spring</th>\n",
              "      <th>course_id_HarvardX/PH207x/2012_Fall</th>\n",
              "      <th>course_id_HarvardX/PH278x/2013_Spring</th>\n",
              "      <th>LoE_DI_Bachelor's</th>\n",
              "      <th>LoE_DI_Doctorate</th>\n",
              "      <th>LoE_DI_Less than Secondary</th>\n",
              "      <th>LoE_DI_Master's</th>\n",
              "      <th>LoE_DI_Secondary</th>\n",
              "      <th>LoE_DI_unk</th>\n",
              "      <th>gender_f</th>\n",
              "      <th>gender_m</th>\n",
              "      <th>gender_unk</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1977</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1979</td>\n",
              "      <td>736</td>\n",
              "      <td>23</td>\n",
              "      <td>0</td>\n",
              "      <td>19</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1984</td>\n",
              "      <td>377</td>\n",
              "      <td>7</td>\n",
              "      <td>49</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1986</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1978</td>\n",
              "      <td>122</td>\n",
              "      <td>5</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   registered  viewed  explored  ...  gender_f  gender_m  gender_unk\n",
              "0           1       0         0  ...         0         1           0\n",
              "1           1       1         1  ...         0         1           0\n",
              "2           1       1         0  ...         1         0           0\n",
              "3           1       0         0  ...         1         0           0\n",
              "4           1       1         0  ...         1         0           0\n",
              "\n",
              "[5 rows x 23 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1257
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "x0fE__dvr_F0"
      },
      "source": [
        "I used the models to predict on the test data and exported and downloaded the files as .csv (I commented out the files.download() line to avoid excessive downloading every time the code is run)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RyrTYOco0gXs"
      },
      "source": [
        "from google.colab import files\n",
        "decision_tree_pred = decision_tree_model.predict(edx_test[features])\n",
        "decision_tree_test_predictions = edx_id.copy() # edx_test[['userid_DI']].copy()\n",
        "decision_tree_test_predictions.loc[:, target] = decision_tree_pred\n",
        "decision_tree_test_predictions.to_csv('submission.csv', index = False)\n",
        "# files.download('submission.csv')\n",
        "\n",
        "small_tree_pred = small_tree_model.predict(edx_test[features])\n",
        "small_tree_test_predictions = edx_id.copy() # edx_test[['userid_DI']].copy()\n",
        "small_tree_test_predictions.loc[:, target] = small_tree_pred\n",
        "small_tree_test_predictions.to_csv('small_tree_submission.csv', index = False)\n",
        "# files.download('small_tree_submission.csv')\n",
        "\n",
        "big_tree_pred = big_tree_model.predict(edx_test[features])\n",
        "big_tree_test_predictions = edx_id.copy() # edx_test[['userid_DI']].copy()\n",
        "big_tree_test_predictions.loc[:, target] = big_tree_pred\n",
        "big_tree_test_predictions.to_csv('big_tree_submission.csv', index = False)\n",
        "# files.download('big_tree_submission.csv')\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EPSqT6pRsdJE"
      },
      "source": [
        "## Grid Search"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XoCGVe3nmJcS"
      },
      "source": [
        "I performed a grid search that resulted in a max depth of 5 and min_samples_leaf of 1 to find the best fit. However, ultimately I decided not to use this information to attempt any more decision trees but rather explore other classification models (below)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D2DJodrgZ7Kw",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "c362c207-07f7-405a-fb22-c803e3fa1503"
      },
      "source": [
        "# TODO\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "hyperparameters = {\n",
        "            'min_samples_leaf': [1, 10, 50, 100, 200, 300],\n",
        "            'max_depth': [1, 5, 10, 15, 20]\n",
        "            }\n",
        "\n",
        "search = GridSearchCV(estimator=DecisionTreeClassifier(random_state=1), param_grid=hyperparameters,cv=6, return_train_score=True)\n",
        "search.fit(train_data[features], train_data[target])\n",
        "search.best_params_"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'max_depth': 10, 'min_samples_leaf': 10}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1259
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZO7wnwHdaYuC",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 411
        },
        "outputId": "ddb89502-533c-48d1-d598-4761506a0a2e"
      },
      "source": [
        "def plot_scores(ax, title, search, hyperparameters, score_key):\n",
        "    cv_results = search.cv_results_\n",
        "    scores = cv_results[score_key]\n",
        "    scores = scores.reshape((len(hyperparameters['max_depth']), len(hyperparameters['min_samples_leaf'])))\n",
        "    max_depths = cv_results['param_max_depth'].reshape(scores.shape).data.astype(int)\n",
        "    min_samples_leafs = cv_results['param_min_samples_leaf'].reshape(scores.shape).data.astype(int)\n",
        "    \n",
        "    ax.plot_wireframe(max_depths, min_samples_leafs, scores)\n",
        "    ax.view_init(20, 220)\n",
        "    ax.set_xlabel('Maximum Depth')\n",
        "    ax.set_ylabel('Minimum Samples Leaf')\n",
        "    ax.set_zlabel('Accuracy')\n",
        "    ax.set_title(title)\n",
        "\n",
        "\n",
        "fig = plt.figure(figsize=(15,7))\n",
        "ax1 = fig.add_subplot(121, projection='3d')\n",
        "ax2 = fig.add_subplot(122, projection='3d')\n",
        "\n",
        "plot_scores(ax1, 'Train Accuracy', search, hyperparameters, 'mean_train_score')\n",
        "plot_scores(ax2, 'Validation Accuracy', search, hyperparameters, 'mean_test_score')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x504 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fpAsXhpiQH-p"
      },
      "source": [
        "## Random Forest\n",
        "For my next three models, I chose Random Forest, AdaBoost, and KNN. After running all three of these models (below) I found that the best validation accuracy out of every model I tried was Random Forest. Afterward, I tuned the features and hyperparameters that resulted in the best possible validation accuracy."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZyhDf1JoQHml",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 158
        },
        "outputId": "85a4d4d7-49c9-4f56-cc32-69e97c9605d0"
      },
      "source": [
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "rfc = RandomForestClassifier(max_depth=25, random_state=1)\n",
        "rfc.fit(train_data[features], train_data[target])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "RandomForestClassifier(bootstrap=True, ccp_alpha=0.0, class_weight=None,\n",
              "                       criterion='gini', max_depth=25, max_features='auto',\n",
              "                       max_leaf_nodes=None, max_samples=None,\n",
              "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                       min_samples_leaf=1, min_samples_split=2,\n",
              "                       min_weight_fraction_leaf=0.0, n_estimators=100,\n",
              "                       n_jobs=None, oob_score=False, random_state=1, verbose=0,\n",
              "                       warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1261
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bXprmHKCRAce",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "95257da5-2a9e-4569-b7c0-c957ef20c8ac"
      },
      "source": [
        "rfc_val_prediction = rfc.predict(validation_data[features])\n",
        "rfc_train_prediction = rfc.predict(train_data[features])\n",
        "rfc_val_acc = accuracy_score(rfc_val_prediction, validation_data[target])\n",
        "rfc_train_acc = accuracy_score(rfc_train_prediction, train_data[target])\n",
        "\n",
        "print(\"random forest classifier validation accuracy:\", rfc_val_acc)\n",
        "print(\"random forest classifier training accuracy:\", rfc_train_acc)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "random forest classifier validation accuracy: 0.976027397260274\n",
            "random forest classifier training accuracy: 0.9998572652012561\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8ds2wKIHR1Y6"
      },
      "source": [
        "rfc_pred = rfc.predict(edx_test[features])\n",
        "rfc_test_predictions = edx_id.copy() # edx_test[['userid_DI']].copy()\n",
        "rfc_test_predictions.loc[:, target] = rfc_pred\n",
        "rfc_test_predictions.to_csv('rfc_submission.csv', index = False)\n",
        "# files.download('rfc_submission.csv')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jK01FVezS2S1"
      },
      "source": [
        "## AdaBoost"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XvUsyturS4If",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "dbc3dd70-c6ea-4a60-b868-9c8853d96c6b"
      },
      "source": [
        "from sklearn.ensemble import AdaBoostClassifier\n",
        "adaboost = AdaBoostClassifier(n_estimators=1000, random_state=1)\n",
        "adaboost.fit(train_data[features], train_data[target])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "AdaBoostClassifier(algorithm='SAMME.R', base_estimator=None, learning_rate=1.0,\n",
              "                   n_estimators=1000, random_state=1)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1264
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TKHiF3tQTOwf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "c0f3813a-1e92-4581-b2be-002929359c4f"
      },
      "source": [
        "adaboost_val_prediction = adaboost.predict(validation_data[features])\n",
        "adaboost_train_prediction = adaboost.predict(train_data[features])\n",
        "adaboost_val_acc = accuracy_score(adaboost_val_prediction, validation_data[target])\n",
        "adaboost_train_acc = accuracy_score(adaboost_train_prediction, train_data[target])\n",
        "\n",
        "print(\"adaboost classifier validation accuracy:\", adaboost_val_acc)\n",
        "print(\"adaboost classifier training accuracy:\", adaboost_train_acc)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "adaboost classifier validation accuracy: 0.9611872146118722\n",
            "adaboost classifier training accuracy: 0.9703111618612618\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hbTszh3WDj38"
      },
      "source": [
        "adaboost_pred = adaboost.predict(edx_test[features])\n",
        "adaboost_test_predictions = edx_id.copy() # edx_test[['userid_DI']].copy()\n",
        "adaboost_test_predictions.loc[:, target] = adaboost_pred\n",
        "adaboost_test_predictions.to_csv('adaboost_submission.csv', index = False)\n",
        "# files.download('adaboost_submission.csv')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DDs7ZbTEUqpN"
      },
      "source": [
        "## K Nearest Neighbors"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MkLtH2GYUt7D",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 70
        },
        "outputId": "59df1166-b6aa-4c30-93fd-519fd1c7d6e0"
      },
      "source": [
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "neigh = KNeighborsClassifier(n_neighbors=4)\n",
        "neigh.fit(train_data[features], train_data[target])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n",
              "                     metric_params=None, n_jobs=None, n_neighbors=4, p=2,\n",
              "                     weights='uniform')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1267
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dwtleN9SU6rB",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "fb013719-e22e-45ae-e302-ae40c3934003"
      },
      "source": [
        "neigh_val_prediction = neigh.predict(validation_data[features])\n",
        "neigh_train_prediction = neigh.predict(train_data[features])\n",
        "neigh_val_acc = accuracy_score(neigh_val_prediction, validation_data[target])\n",
        "neigh_train_acc = accuracy_score(neigh_train_prediction, train_data[target])\n",
        "\n",
        "print(\"knn classifier validation accuracy:\", neigh_val_acc)\n",
        "print(\"knn classifier training accuracy:\", neigh_train_acc)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "knn classifier validation accuracy: 0.966324200913242\n",
            "knn classifier training accuracy: 0.9701684270625178\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QRaqxxYpD1o2"
      },
      "source": [
        "neigh_pred = neigh.predict(edx_test[features])\n",
        "neigh_test_predictions = edx_id.copy() # edx_test[['userid_DI']].copy()\n",
        "neigh_test_predictions.loc[:, target] = neigh_pred\n",
        "neigh_test_predictions.to_csv('neigh_submission.csv', index = False)\n",
        "# files.download('neigh_submission.csv')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-J5gUFrIof7V"
      },
      "source": [
        "# Plot ROC Curves"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Z0GPPTn1senI",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 932
        },
        "outputId": "0788e5d2-3dc1-49e1-9fdc-e33dd11fcc76"
      },
      "source": [
        "from sklearn.metrics import roc_curve\n",
        "from matplotlib import pyplot\n",
        "import os\n",
        "\n",
        "# get probabilities\n",
        "ns_probs = [0 for _ in range(len(validation_data))]\n",
        "decision_tree_probs = decision_tree_model.predict_proba(validation_data[features])\n",
        "decision_tree_probs = decision_tree_probs[:, 1]\n",
        "\n",
        "small_tree_probs = small_tree_model.predict_proba(validation_data[features])\n",
        "small_tree_probs = small_tree_probs[:, 1]\n",
        "\n",
        "big_tree_probs = big_tree_model.predict_proba(validation_data[features])\n",
        "big_tree_probs = big_tree_probs[:, 1]\n",
        "\n",
        "rfc_probs = rfc.predict_proba(validation_data[features])\n",
        "rfc_probs = rfc_probs[:, 1]\n",
        "\n",
        "adaboost_probs = adaboost.predict_proba(validation_data[features])\n",
        "adaboost_probs = adaboost_probs[:, 1]\n",
        "\n",
        "knn_probs = neigh.predict_proba(validation_data[features])\n",
        "knn_probs = knn_probs[:, 1]\n",
        "\n",
        "# get roc curves\n",
        "ns_fpr, ns_tpr, _ = roc_curve(validation_data[target], ns_probs)\n",
        "decision_tree_fpr, decision_tree_tpr, _ = roc_curve(validation_data[target], decision_tree_probs)\n",
        "small_tree_fpr, small_tree_tpr, _ = roc_curve(validation_data[target], small_tree_probs)\n",
        "big_tree_fpr, big_tree_tpr, _ = roc_curve(validation_data[target], big_tree_probs)\n",
        "\n",
        "rfc_fpr, rfc_tpr, _ = roc_curve(validation_data[target], rfc_probs)\n",
        "adaboost_fpr, adaboost_tpr, _ = roc_curve(validation_data[target], adaboost_probs)\n",
        "knn_fpr, knn_tpr, _ = roc_curve(validation_data[target], knn_probs)\n",
        "\n",
        "\n",
        "plt.figure(figsize=(15, 15))\n",
        "\n",
        "# plot the roc curve for the model\n",
        "pyplot.plot(ns_fpr, ns_tpr, linestyle='--', label='No Skill')\n",
        "pyplot.plot(decision_tree_fpr, decision_tree_tpr, marker='.', label='Decision Tree, max depth = 5')\n",
        "pyplot.plot(small_tree_fpr, small_tree_tpr, marker='.', label='Small Decision Tree, max depth = 2')\n",
        "# pyplot.plot(big_tree_fpr, big_tree_tpr, marker='.', label='Big Decision Tree')\n",
        "\n",
        "pyplot.plot(rfc_fpr, rfc_tpr, marker='.', label='Random Forest')\n",
        "pyplot.plot(adaboost_fpr, adaboost_tpr, marker='.', label='AdaBoost')\n",
        "pyplot.plot(knn_fpr, knn_tpr, marker='.', label='K Nearest Neighbors')\n",
        "pyplot.title(\"ROC Curves\", fontsize=20)\n",
        "# axis labels\n",
        "pyplot.xlabel('False Positive Rate', fontsize=20)\n",
        "pyplot.ylabel('True Positive Rate', fontsize=20)\n",
        "pyplot.xticks(fontsize=15)\n",
        "pyplot.yticks(fontsize=15)\n",
        "# show the legend\n",
        "pyplot.legend()\n",
        "\n",
        "filename=\"ROC-curves.png\"\n",
        "pyplot.savefig(filename)\n",
        "print(\"Image file saved to \"+os.getcwd()+\"/\"+filename)\n",
        "\n",
        "# show the plot\n",
        "pyplot.show()\n",
        "\n",
        "from google.colab import files\n",
        "# files.download(filename)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Image file saved to /content/ROC-curves.png\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x1080 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}