digit_recognition.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "# 机器学习工程师纳米学位\n",
    "## 深度学习\n",
    "## 项目：搭建一个数字识别项目\n",
    "\n",
    "在此文件中，我们提供给你了一个模板，以便于你根据项目的要求一步步实现要求的功能，进而完成整个项目。如果你认为需要导入另外的一些代码，请确保你正确导入了他们，并且包含在你的提交文件中。以**'练习'**开始的标题表示接下来你将开始实现你的项目。注意有一些练习是可选的，并且用**'可选'**标记出来了。\n",
    "\n",
    "在此文件中，有些示例代码已经提供给你，但你还需要实现更多的功能让项目成功运行。除非有明确要求，你无须修改任何已给出的代码。以'练习'开始的标题表示接下来的代码部分中有你必须要实现的功能。每一部分都会有详细的指导，需要实现的部分也会在注释中以'TODO'标出。请仔细阅读所有的提示！\n",
    "\n",
    "除了实现代码外，你还必须回答一些与项目和你的实现有关的问题。每一个需要你回答的问题都会以**'问题 X'**为标题。请仔细阅读每个问题，并且在问题后的**'回答'**文字框中写出完整的答案。我们将根据你对问题的回答和撰写代码所实现的功能来对你提交的项目进行评分。\n",
    "\n",
    ">**注意：** Code 和 Markdown 区域可通过 **Shift + Enter** 快捷键运行。此外，Markdown可以通过双击进入编辑模式。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "## 连接 mnist 的字符来合成数据\n",
    "\n",
    "你可以通过连接[MNIST](http://yann.lecun.com/exdb/mnist/)的字符来合成数据来训练这个模型。为了快速导入数据集，我们可以使用 [Keras Datasets](https://keras.io/datasets/#mnist-database-of-handwritten-digits) [中文文档](http://keras-cn.readthedocs.io/en/latest/other/datasets/#mnist)。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 载入 mnist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://s3.amazonaws.com/img-datasets/mnist.pkl.gz\n"
     ]
    }
   ],
   "source": [
    "from keras.datasets import mnist\n",
    "\n",
    "(X_raw, y_raw), (X_raw_test, y_raw_test) = mnist.load_data()\n",
    "\n",
    "n_train, n_test = X_raw.shape[0], X_raw_test.shape[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 可视化 mnist\n",
    "\n",
    "我们可以通过 matplotlib 来可视化我们的原始数据集。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAuUAAAIDCAYAAACjJ2BhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYAUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fbbe7dc6e48>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 257,
       "width": 370
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import random\n",
    "\n",
    "%matplotlib inline\n",
    "%config InlineBackend.figure_format = 'retina'\n",
    "\n",
    "for i in range(15):\n",
    "    plt.subplot(3, 5, i+1)\n",
    "    index = random.randint(0, n_train-1)\n",
    "    plt.title(str(y_raw[index]))\n",
    "    plt.imshow(X_raw[index], cmap='gray')\n",
    "    plt.axis('off')\n",
    "\n",
    "# print X_raw.shape\n",
    "# print y_raw.shape\n",
    "# print X_raw_test.shape\n",
    "# print y_raw_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 练习：合成数据\n",
    "\n",
    "你需要随机取随机张图片，然后将它们拼接成新的图片。\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "\n",
    "n_class, n_len, width, height = 11, 5, 28, 28\n",
    "\n",
    "def generate_dataset(X, y):\n",
    "    X_len = X.shape[0]\n",
    "    \n",
    "    X_gen = np.zeros((X_len, height, width*n_len, 1), dtype=np.uint8)\n",
    "    y_gen = [np.zeros((X_len, n_class), dtype=np.uint8) for i in range(n_len)]\n",
    "    # TODO: 随机取1~5个数字，并拼接成新的图片\n",
    "    for index in range(X_len):        \n",
    "        digits = random.randint(1, 5)    # No. of number to select\n",
    "        # use slicing instead of stack\n",
    "        for i in range(digits):\n",
    "            j = random.randint(0, X_len-1)   # select number randomly by index 'j'\n",
    "            X_gen[index,:,i*width:(i+1)*width,0] = X[j]\n",
    "            y_gen[i][index][y[j]] = 1\n",
    "        for i in range(digits, n_len):\n",
    "            y_gen[i][index][10] = 1\n",
    "#         length = random.randint(1, 5)\n",
    "#         lst_digit = random.sample(range(X_len), length)                        \n",
    "#         mat_pic = np.zeros((28, 28 * (n_len-length)))    # fill the empty part with 0-matrix\n",
    "#         for i in lst_digit[::-1]:\n",
    "#             mat_pic = np.hstack((X[i], mat_pic))  # insert the valid digit in head                        \n",
    "#         X_gen[index][:,:,0] = mat_pic\n",
    "        # label\n",
    "#         for i in range(5):\n",
    "#             if i < len(lst_digit):\n",
    "#                 y_gen[i][index][y[lst_digit[i]]] = 1   # valid digit (0~9)\n",
    "#             else:\n",
    "#                 y_gen[i][index][10] = 1          # 10 stands for empty\n",
    "    return X_gen, y_gen\n",
    "\n",
    "X_train, y_train = generate_dataset(X_raw, y_raw)\n",
    "X_test, y_test = generate_dataset(X_raw_test, y_raw_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAuUAAAHsCAYAAAB13dOrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n6lGjpcjnw7IAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fbbe1ad4710>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 246,
       "width": 370
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 显示生成的图片\n",
    "for i in range(15):\n",
    "    plt.subplot(5, 3, i+1)\n",
    "    index = random.randint(0, n_test-1)\n",
    "    title = ''\n",
    "    for j in range(n_len):\n",
    "        title += str(np.argmax(y_test[j][index])) + ','\n",
    "    \n",
    "    plt.title(title)\n",
    "    plt.imshow(X_test[index][:,:,0], cmap='gray')\n",
    "    plt.axis('off')\n",
    "\n",
    "# print X_train.shape\n",
    "# print len(y_train)\n",
    "# print X_test.shape\n",
    "# print len(y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 问题 1\n",
    "_你是如何合成数据集的？_\n",
    "\n",
    "**回答：** \n",
    "* ~~由于根据现有代码，最终返回的应该是原始输入数目的数据集，只是每一项的图片尺寸曾大为原始的五倍（图片拼接），~~\n",
    "* ~~所以，对每一个输入，首先我从1~5中随机抽取一个数字，然后从原始的输入中随机抽取该数字数目的下标值并保存；~~\n",
    "* ~~接下来由于每一个单位的输入为原始的5倍，所以根据前面的抽取的数据长度，我先初始化需要补充的数目的图片（像素点）矩阵~~\n",
    "* ~~然后逐个将所抽取的数字下标值对应的图片矩阵插入到初始化的补充图片矩阵的头部，构成一个单位的输入图片矩阵~~\n",
    "* ~~最后就是对应的讲所构成的图片矩阵加入每一个对应的输入区域，对于label y则填充原始输入中对应的下标值，其中对于空缺位填充数字值为10~~\n",
    "---\n",
    "由于本项目最终要求是构建一个对数字图片序列进行识别的模型，所以，需要将原始训练集和测试集的每一个数字与其他随机取出的不超过4个数字进行拼接，形成一个不超过5个数字的序列\n",
    "* 对传入的数据集的每一个数字，先随机一个不超过5的数字作为这一数字序列的长度\n",
    "* 针对上一步随机生成的数值位数的每一位，从原始数据集随机提取一个数字对应的下标，然后利用slicing将X中的该图片对应的矩阵填充到这一数位对应的数字中，对应的label标记为1\n",
    "* 最后针对不足5位的X值（空图片）对应的label填充数字为10"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 练习：设计并测试一个模型架构\n",
    "\n",
    "设计并实现一个能够识别数字序列的深度学习模型。为了产生用于测试的合成数字序列，你可以进行如下的设置：比如，你可以限制一个数据序列最多五个数字，并在你的深度网络上使用五个分类器。同时，你有必要准备一个额外的“空白”的字符，以处理相对较短的数字序列。\n",
    "\n",
    "在思考这个问题的时候有很多方面可以考虑：\n",
    "\n",
    "- 你的模型可以基于深度神经网络或者是卷积神经网络。\n",
    "- 你可以尝试是否在每个分类器间共享权值。\n",
    "- 你还可以在深度神经网络中使用循环网络来替换其中的分类层，并且将数字序列里的数字一个一个地输出。\n",
    "\n",
    "在使用 Keras 搭建模型的时候，你可以使用 [泛型模型](http://keras-cn.readthedocs.io/en/latest/models/model/) 的方式来搭建多输出模型。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "from keras.models import Model\n",
    "from keras.layers import *\n",
    "\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Flatten\n",
    "from keras.layers.convolutional import Convolution2D, MaxPooling2D\n",
    "from keras.utils import np_utils\n",
    "\n",
    "# TODO: 构建你的模型\n",
    "# fix random seed for reproducibility\n",
    "seed = 7\n",
    "np.random.seed(seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "# # normalize inputs from 0-255 to 0-1\n",
    "# X_train /= 255\n",
    "# X_test /= 255\n",
    "# # one hot encode outputs\n",
    "# y_train = np_utils.to_categorical(y_train)\n",
    "# y_test = np_utils.to_categorical(y_test)\n",
    "# num_classes = y_test.shape[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "def train_model():\n",
    "    # create model\n",
    "#     model = Sequential()\n",
    "    a = Input((28, 28*5, 1))\n",
    "#     b = Convolution2D(30, 5, 5, border_mode='same', activation='relu')(a)\n",
    "    b = Convolution2D(30, 3, 3, border_mode='same', activation='relu')(a)\n",
    "    b = Convolution2D(30, 3, 3, border_mode='same', activation='relu')(b)\n",
    "    b = MaxPooling2D(pool_size=(2, 2), border_mode='same')(b)\n",
    "    b = Convolution2D(15, 3, 3, border_mode='same', activation='relu')(b)\n",
    "    b = Convolution2D(15, 3, 3, border_mode='same', activation='relu')(b)\n",
    "    b = MaxPooling2D(pool_size=(2, 2),border_mode='same')(b)    \n",
    "    b = Flatten()(b)\n",
    "    b = Dropout(0.2)(b)\n",
    "#     b = Dense(128, activation='relu')(b)\n",
    "#     b = Dense(50, activation='relu')(b)\n",
    "#     model.add(Convolution2D(30, 5, 5, border_mode='valid', input_shape=(28, 28*5, 1), activation='relu'))\n",
    "#     model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "#     model.add(Convolution2D(15, 3, 3, activation='relu'))\n",
    "#     model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "#     model.add(Dropout(0.2))\n",
    "#     model.add(Flatten())\n",
    "#     model.add(Dense(128, activation='relu'))\n",
    "#     model.add(Dense(50, activation='relu'))\n",
    "#     lst_output = []\n",
    "#     for i in range(5):\n",
    "#         lst_output.append(Dense(11, activation='softmax')(b))\n",
    "    # use list comprehension to replace the loop\n",
    "    lst_output = [Dense(11, activation='softmax')(b) for i in range(n_len)]\n",
    "    model = Model(input=a, output=lst_output)\n",
    "    # compile model\n",
    "    model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 问题 2\n",
    "_你为解决这个问题采取了什么技术？请详细介绍你使用的技术。_\n",
    "\n",
    "**回答：** \n",
    "* 卷积层:用来提取输入的图片的特征，将原始数据映射到隐藏层的特征空间\n",
    "* 使用ReLU作为激活函数:用来增强判定函数和整个神经网络的非线性特性，相较于其他的激活函数而言，ReLU求导速度快，所以可以将神经网络的训练速度提升,~~而不会对模型的泛化造成显著影响~~此外，ReLU解决了Sigmoid激活函数会出现的梯度消失的问题\n",
    "* Flatten：通过将二位的矩阵数据转换为向量，使得后续的数据可以被全连接层更好的处理\n",
    "* Dropout:随机的抛弃该层的20%的神经元来防止过拟合\n",
    "* softmax:~~对于每一类以概率的形式输出概率最大的预测结果(Max-layer output)~~针对每一个数字输出对应的概率，而且针对其中概率最大的数值，相对于其他的数值，softmax会增大其概率，并且在此基础之上保证所有的数值概率之和为1\n",
    "* 最大池化：将输入的图像划分为若干个2\\*2的矩形区域后对每个矩形区域输出最大值，用来减小数据的空间大小，此外一定程度上也防止了过拟合\n",
    "* 使用Adam作为优化器的优化算法，因为该算法计算效率较高"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 可视化你的网络模型\n",
    "\n",
    "参考链接：[visualization](http://keras-cn.readthedocs.io/en/latest/other/visualization/)\n",
    "\n",
    "可以是 PNG 格式，也可以是 SVG 格式。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABwEAAAQtCAIAAAA9bacDAAAABmJLR0QA/wD/AP+gvaeTAAAgAElE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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# from keras.utils.visualize_util import plot, model_to_dot\n",
    "# from keras.utils.visualize_util import plot\n",
    "from IPython.display import Image, SVG\n",
    "# from IPython.display import Image\n",
    "\n",
    "# TODO: 可视化你的模型\n",
    "model = train_model()\n",
    "# plot(model, to_file='train_model.png', show_shapes=True)\n",
    "Image(filename = 'train_model.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 问题 3\n",
    "_你最终的模型架构是什么样的？（什么类型的模型，层数，大小, 如何连接等）_\n",
    "\n",
    "**回答：**\n",
    "如图所示，我采用了：\n",
    "* Sequential模型\n",
    "* 不含输入层的话一共有~~七~~九层，其中\n",
    " * 第一层为30个卷积核，3行3列的卷积层，直接连接输入层和下一层\n",
    " * 第二层为30个卷积核，3行3列的卷积层，直接连接上一个卷积层和下一层\n",
    " * 第三层为2\\*2的最大池化层，直接与上一层卷积层相连\n",
    " * 第四层为15个卷积核，3行3列的卷积层，同样是直接与上一层相连\n",
    " * 第五层为15个卷积核，3行3列的卷积层，同样是直接与上一层相连\n",
    " * 第六层为2\\*2的最大池化层，直接与上一层卷积层相连\n",
    " * 第七层为Flatten层\n",
    " * 第八层为Dropout层\n",
    " * 最后第九层为5个全连接层，与上一层直接相连，作为输出层"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 练习：训练你的网络模型\n",
    "\n",
    "你需要设置20%的数据作为验证集，以保证模型没有过拟合。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 48000 samples, validate on 12000 samples\n",
      "Epoch 1/25\n",
      "124s - loss: 3.6556 - dense_1_loss: 0.9365 - dense_2_loss: 0.9019 - dense_3_loss: 0.7548 - dense_4_loss: 0.6098 - dense_5_loss: 0.4526 - dense_1_acc: 0.7524 - dense_2_acc: 0.7797 - dense_3_acc: 0.8295 - dense_4_acc: 0.8688 - dense_5_acc: 0.9226 - val_loss: 0.5371 - val_dense_1_loss: 0.1581 - val_dense_2_loss: 0.1512 - val_dense_3_loss: 0.1001 - val_dense_4_loss: 0.0779 - val_dense_5_loss: 0.0499 - val_dense_1_acc: 0.9530 - val_dense_2_acc: 0.9522 - val_dense_3_acc: 0.9684 - val_dense_4_acc: 0.9744 - val_dense_5_acc: 0.9848\n",
      "Epoch 2/25\n",
      "29s - loss: 0.4792 - dense_1_loss: 0.1488 - dense_2_loss: 0.1215 - dense_3_loss: 0.0947 - dense_4_loss: 0.0753 - dense_5_loss: 0.0389 - dense_1_acc: 0.9531 - dense_2_acc: 0.9610 - dense_3_acc: 0.9704 - dense_4_acc: 0.9765 - dense_5_acc: 0.9875 - val_loss: 0.3514 - val_dense_1_loss: 0.1077 - val_dense_2_loss: 0.0884 - val_dense_3_loss: 0.0626 - val_dense_4_loss: 0.0530 - val_dense_5_loss: 0.0398 - val_dense_1_acc: 0.9676 - val_dense_2_acc: 0.9723 - val_dense_3_acc: 0.9809 - val_dense_4_acc: 0.9825 - val_dense_5_acc: 0.9878\n",
      "Epoch 3/25\n",
      "29s - loss: 0.3108 - dense_1_loss: 0.1014 - dense_2_loss: 0.0813 - dense_3_loss: 0.0590 - dense_4_loss: 0.0478 - dense_5_loss: 0.0214 - dense_1_acc: 0.9679 - dense_2_acc: 0.9741 - dense_3_acc: 0.9812 - dense_4_acc: 0.9850 - dense_5_acc: 0.9927 - val_loss: 0.2822 - val_dense_1_loss: 0.0824 - val_dense_2_loss: 0.0677 - val_dense_3_loss: 0.0545 - val_dense_4_loss: 0.0379 - val_dense_5_loss: 0.0396 - val_dense_1_acc: 0.9747 - val_dense_2_acc: 0.9795 - val_dense_3_acc: 0.9810 - val_dense_4_acc: 0.9873 - val_dense_5_acc: 0.9875\n",
      "Epoch 4/25\n",
      "29s - loss: 0.2277 - dense_1_loss: 0.0760 - dense_2_loss: 0.0593 - dense_3_loss: 0.0428 - dense_4_loss: 0.0344 - dense_5_loss: 0.0152 - dense_1_acc: 0.9746 - dense_2_acc: 0.9809 - dense_3_acc: 0.9859 - dense_4_acc: 0.9888 - dense_5_acc: 0.9952 - val_loss: 0.2471 - val_dense_1_loss: 0.0682 - val_dense_2_loss: 0.0580 - val_dense_3_loss: 0.0518 - val_dense_4_loss: 0.0338 - val_dense_5_loss: 0.0353 - val_dense_1_acc: 0.9793 - val_dense_2_acc: 0.9830 - val_dense_3_acc: 0.9838 - val_dense_4_acc: 0.9892 - val_dense_5_acc: 0.9902\n",
      "Epoch 5/25\n",
      "29s - loss: 0.1814 - dense_1_loss: 0.0629 - dense_2_loss: 0.0493 - dense_3_loss: 0.0337 - dense_4_loss: 0.0245 - dense_5_loss: 0.0110 - dense_1_acc: 0.9803 - dense_2_acc: 0.9843 - dense_3_acc: 0.9891 - dense_4_acc: 0.9916 - dense_5_acc: 0.9963 - val_loss: 0.2067 - val_dense_1_loss: 0.0592 - val_dense_2_loss: 0.0515 - val_dense_3_loss: 0.0353 - val_dense_4_loss: 0.0306 - val_dense_5_loss: 0.0300 - val_dense_1_acc: 0.9818 - val_dense_2_acc: 0.9838 - val_dense_3_acc: 0.9890 - val_dense_4_acc: 0.9905 - val_dense_5_acc: 0.9921\n",
      "Epoch 6/25\n",
      "29s - loss: 0.1424 - dense_1_loss: 0.0500 - dense_2_loss: 0.0407 - dense_3_loss: 0.0262 - dense_4_loss: 0.0185 - dense_5_loss: 0.0070 - dense_1_acc: 0.9828 - dense_2_acc: 0.9865 - dense_3_acc: 0.9911 - dense_4_acc: 0.9937 - dense_5_acc: 0.9974 - val_loss: 0.2070 - val_dense_1_loss: 0.0561 - val_dense_2_loss: 0.0523 - val_dense_3_loss: 0.0368 - val_dense_4_loss: 0.0279 - val_dense_5_loss: 0.0338 - val_dense_1_acc: 0.9826 - val_dense_2_acc: 0.9842 - val_dense_3_acc: 0.9883 - val_dense_4_acc: 0.9918 - val_dense_5_acc: 0.9918\n",
      "Epoch 7/25\n",
      "29s - loss: 0.1228 - dense_1_loss: 0.0425 - dense_2_loss: 0.0342 - dense_3_loss: 0.0241 - dense_4_loss: 0.0153 - dense_5_loss: 0.0067 - dense_1_acc: 0.9850 - dense_2_acc: 0.9889 - dense_3_acc: 0.9920 - dense_4_acc: 0.9945 - dense_5_acc: 0.9979 - val_loss: 0.2018 - val_dense_1_loss: 0.0591 - val_dense_2_loss: 0.0519 - val_dense_3_loss: 0.0361 - val_dense_4_loss: 0.0266 - val_dense_5_loss: 0.0283 - val_dense_1_acc: 0.9816 - val_dense_2_acc: 0.9839 - val_dense_3_acc: 0.9882 - val_dense_4_acc: 0.9915 - val_dense_5_acc: 0.9925\n",
      "Epoch 8/25\n",
      "29s - loss: 0.1046 - dense_1_loss: 0.0362 - dense_2_loss: 0.0291 - dense_3_loss: 0.0197 - dense_4_loss: 0.0148 - dense_5_loss: 0.0048 - dense_1_acc: 0.9883 - dense_2_acc: 0.9899 - dense_3_acc: 0.9932 - dense_4_acc: 0.9947 - dense_5_acc: 0.9984 - val_loss: 0.1965 - val_dense_1_loss: 0.0519 - val_dense_2_loss: 0.0519 - val_dense_3_loss: 0.0342 - val_dense_4_loss: 0.0256 - val_dense_5_loss: 0.0329 - val_dense_1_acc: 0.9846 - val_dense_2_acc: 0.9837 - val_dense_3_acc: 0.9893 - val_dense_4_acc: 0.9928 - val_dense_5_acc: 0.9916\n",
      "Epoch 9/25\n",
      "29s - loss: 0.0891 - dense_1_loss: 0.0308 - dense_2_loss: 0.0264 - dense_3_loss: 0.0159 - dense_4_loss: 0.0125 - dense_5_loss: 0.0034 - dense_1_acc: 0.9897 - dense_2_acc: 0.9911 - dense_3_acc: 0.9947 - dense_4_acc: 0.9956 - dense_5_acc: 0.9988 - val_loss: 0.2154 - val_dense_1_loss: 0.0636 - val_dense_2_loss: 0.0522 - val_dense_3_loss: 0.0353 - val_dense_4_loss: 0.0307 - val_dense_5_loss: 0.0335 - val_dense_1_acc: 0.9811 - val_dense_2_acc: 0.9856 - val_dense_3_acc: 0.9896 - val_dense_4_acc: 0.9909 - val_dense_5_acc: 0.9922\n",
      "Epoch 10/25\n",
      "29s - loss: 0.0831 - dense_1_loss: 0.0309 - dense_2_loss: 0.0228 - dense_3_loss: 0.0150 - dense_4_loss: 0.0106 - dense_5_loss: 0.0039 - dense_1_acc: 0.9896 - dense_2_acc: 0.9923 - dense_3_acc: 0.9948 - dense_4_acc: 0.9965 - dense_5_acc: 0.9985 - val_loss: 0.2080 - val_dense_1_loss: 0.0575 - val_dense_2_loss: 0.0471 - val_dense_3_loss: 0.0346 - val_dense_4_loss: 0.0281 - val_dense_5_loss: 0.0408 - val_dense_1_acc: 0.9823 - val_dense_2_acc: 0.9876 - val_dense_3_acc: 0.9890 - val_dense_4_acc: 0.9916 - val_dense_5_acc: 0.9909\n",
      "Epoch 11/25\n",
      "29s - loss: 0.0747 - dense_1_loss: 0.0281 - dense_2_loss: 0.0202 - dense_3_loss: 0.0127 - dense_4_loss: 0.0097 - dense_5_loss: 0.0039 - dense_1_acc: 0.9906 - dense_2_acc: 0.9933 - dense_3_acc: 0.9954 - dense_4_acc: 0.9966 - dense_5_acc: 0.9987 - val_loss: 0.1967 - val_dense_1_loss: 0.0493 - val_dense_2_loss: 0.0499 - val_dense_3_loss: 0.0380 - val_dense_4_loss: 0.0256 - val_dense_5_loss: 0.0339 - val_dense_1_acc: 0.9847 - val_dense_2_acc: 0.9845 - val_dense_3_acc: 0.9884 - val_dense_4_acc: 0.9916 - val_dense_5_acc: 0.9922\n"
     ]
    }
   ],
   "source": [
    "# TODO: 训练你的模型\n",
    "# log = model.fit(X_train, [label for label in y_train], nb_epoch=10, validation_split=0.2, batch_size=200, verbose=2)\n",
    "from keras.callbacks import EarlyStopping\n",
    "\n",
    "early_stop = EarlyStopping(monitor='val_loss', min_delta=0.0005, patience=2)\n",
    "log = model.fit(X_train, [label for label in y_train], nb_epoch=25, validation_split=0.2, batch_size=200, verbose=2,\n",
    "               callbacks=[early_stop])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 练习：计算你的模型准确率\n",
    "\n",
    "我们刚才得到了模型每个数字的准确率，现在让我们来计算整体准确率，按照完全预测正确数字序列的标准来计算。\n",
    "\n",
    "比如 1,2,3,10,10 预测成了 1,2,10,10,10 算错，而不是算对了80%。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9359\n"
     ]
    }
   ],
   "source": [
    "def evaluate(model):\n",
    "    # TODO: 按照错一个就算错的规则计算准确率\n",
    "    y_pred = model.predict(X_test, batch_size=200)\n",
    "    pred = np.argmax(y_pred, axis=2)\n",
    "    test = np.argmax(y_test, axis=2)\n",
    "    accuracy = np.min((pred==test), axis=0).mean()\n",
    "    return accuracy\n",
    "\n",
    "print(evaluate(model))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 问题 4\n",
    "\n",
    "_你的模型准确率有多少？你觉得你的模型足以解决问题吗？_\n",
    "\n",
    "**回答：**\n",
    "最终模型的准确率为~~0.9338~~0.9359，在原始模型的基础上替换一层卷积层为3\\*3并额外增加了两层卷积层，此外在训练数据集时添加了callback中的earlyStopping操作是模型在针对验证集没有明显的改善时结束训练，针对该数据应该是有不错的效果，但是如果放到现实中的图片识别，恐怕还需要对数据进行更多的预处理以及模型优化操作"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 预测值可视化\n",
    "\n",
    "我们将模型的预测结果和真实值画出来，观察真实效果。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.5, 139.5, 27.5, -0.5)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAuUAAAD+CAYAAACdtHuPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAAWJQAAFiUBSVIk8AAAIABJREFUeJzt3Xm4HFW57/HfLwlDSEKCDAZFDYIeBBREEUSGiKLoFTmi\nIChHA4qCF0UGCYh4QPEoIuDAFQTF4AVE8IgIKohoVBSQmcMkogkyBIIoUwLEkPf8sVaxe3d19967\nd3avneT7eZ56OrVqraq3q/Ikb61etcoRIQAAAADljCkdAAAAALCiIykHAAAACiMpBwAAAAojKQcA\nAAAKIykHAAAACiMpBwAAAAojKQcAAAAKIykHAAAACiMpBwAAAAojKQcAAAAKIykHAAAACiMpBwAA\nAAojKQcAAAAKIykHsFyyPcN22J5dOhYAAAZCUg4AQ2R7dk74Oy2ndLnvl9ne1/Y3bV9r+5m8v6uH\nsI+ptr9m+y+2n7b9kO2Lbb+pm5jyPifZfqftz9v+ue2/N3zXjQa5jzG2P2L7KtuP2n7C9o22P2V7\n5WHEtrXtg2yfbftO20tyXF8awj5ea/s82w/kc/Y329+2veEw4lrP9vttf9X2720vyHE9OIR9rG77\nONt32F5o+xHbV9h+T7dxARidxpUOAACWYY9LeqrDtm6cIGnXLtvK9qsk/UrSmg1xrCXpHZL+j+1P\nR8Sgk9UGb5J04TDiWknSjyW9PRctkvSspM3zsrvtHSPiyS52f6mkycOI7YOSvq30f2IonbMXSfqQ\npD1tvzMiftXFrg+TdNAw4lpP0m8lrZ+LnpS0uqQdJe1o+9SI+Fi3+wcwutBTDgDdOygiprZZPt3l\nPp+VdIek70n6hKT/P9iGtsdL+olSQn6jpE0jYrKkNSSdKMmS/sv2W7qMbb6kn0k6VtJHhtj2OKWE\n/GlJMyStJmmCpF0k/UPSlpK+1WVcT0n6o6T/J2kfSTcNtmG+iTlDKSE/R9LzI2KKpGmSLs8x/rft\ntbuIKyT9RdIPlBL0k4YQlyX9UCkhnyvpDRExSdIkSYdLWiLpANv7dREXgFGInnIAGF32iIhnqxXb\nxwyh7UclvUSpR3WXiLhfkiLicUmH2d5A0r9L+qKkXwwxrosj4scNcU0bbEPbU9XXYzwzIs5q2HyJ\n7X2VetH3sn18RNwyxNjWazpnM4bQ9nOSVpJ0naQPVvuJiHts7ybpdqVe8yMkHTrEuA6LiIO7jGtX\nSVspJd/vioibclxPSzrB9gskfVLS52yfFRGLhhgbgFGGnnIAXbM9N4+RnW77hXkc9F/zOOhab6Xt\nbfO43ftynUds/9L2XrlnsNUxtrD9JdtX5nG+VbvZtj9se+zIf9PeaUwuu/D+/HlulZA3OSF/bmH7\n33oY17slrSLpMUmnt9j3RZLuUurJf99Qd95tbLanqG84zUnN+8lDaU7Lq23/ji7tuLLqWv6ySsib\nfEWpJ36q0nAWAMs4knIAS8PLlYYMHCDp+ZL+1VzB9vGSfifpvZJeqDSMYQ2lscrnSjrXdqt/k34h\naaakNygNy1go6XmSdlAadvAT20P61S/fRFQPKU4fStvRyvYkSa/Jq5e1qXa1UmIspfPeK2/Mn7/N\nPb2tVD33vUwwt1XqJW88frPqXK4r6RUjHlGf6py1vJb5puu2vEpSDiwHSMoBLA0nSpqnNO51QkRM\nlPTc7BC2D1IaB/uQ0ljkKXms8wRJe0p6MH/ObLHvX0jaS9K6ed9rSJoo6T9yu7dLOrhFu144LM/W\nscj2w3lWjANsr1ogllco9TRLfclaPxGxRNKf8urGvQiq6Vgt48puz5+vGGqP9DBUcT0YEY+0qXN7\nw597cs5sr6O+B3UHc856eS0BjBCScgBLw2JJO0XEH6qCiLhbem6IwHFKPeNvjYgzIuKxXOepiPiB\npN2UfoqvTY0XEe+LiPMi4sGGsgURcbakPXJRqRkoNlHqtV+gNMPJjpK+KemPtl/c41jWbfjzAx3q\nVdvW7VBnaauONZi4JualFwaMKyKekvRoU/2RNpqvJYARQlIOYGn4XkQ81Gbbu5WSrF9GxM2tKkTE\nVZLmKA1neU2rOm3a/U4pYZqWH3wbbLvZEeG8zB5suwazJX1AKRkan3vvny/p05KekfRKST8bztzb\nXZjQ8Od20zRKafiP1LvEV+qLbTBxSb2LbTBxSb0/Z6P5WgIYIcy+AmBpuKrDtm3y544DvDTlefnz\nRc37s7270oNvW0haW1Kr4SEvUOdexaUmIo5pUTZf0hdt3yLpEqVe9Blq8WAjAADNSMoBLA0Pd9hW\n/bS+Wl4G8lyd/ADn+ZLe1bD9GUl/V5rPW0pJ+hj1710sJiJ+avu3krZXmoO7V0n5goY/j5f0RJt6\n1fnt5iU93VogaYpSXO00/t3oVWzVOesUl9T7c9Z8LdspcS0BjBCGrwBYGjpN/Vb9O/O1hiEjnZZZ\nDW33U0rIFyq9SOdFEbFqRKxdvaRHfb3jvXo4cDCuyZ8v7eExG38l6DSUp9o2bwRjaVbFNpi4noyI\ndjcUS9uAceUXMk3Jq706Z6P5WgIYISTlAEZaNda8mwcfd8+fn4+Ib0TEfY0b8xzlaw0nuOXInUoP\ny0pp6ExNnnKymp/89lZ1Rkh1rJZxZdUMIneMcCyNqrim2l6zTZ3GmU16cs4i4mGlX4OkwZ2zXl5L\nACOEpBzASKvGh0/PvY5DsV7+vLHN9jeo9fjy0rbKn3N6dcDcu3xdXt2pTbWtJE3Of75ixIPq8+v8\nuV2H6SKrmHsZ15Xqm1P/zW3qvCV/PqDe3jBU56zltbT9QvUl7L08ZwBGCEk5gJF2gdIY2TUkfbZT\nRdtrNBVVL7p5ZYu645SmWuypgebQtr2z0nhySfrpyEfUz7n58/22W02Td1j+vD4i/tRi+0j5kdKz\nAFMkfbh5o+1dlHrwQ9L3exVUnprzZ3n1kOaXV9meIGn/vPr9iAj1TnUt32J7sxbbD1EasjVPfQk8\ngGUYSTmAEZVfynJkXj3C9hm2X15ttz3e9na2T5X0h6bml+fPo23vmoeryPZGki6W9Dr1fyhuUIb5\nRs8jbJ9pe6f8Fs1qn2vbPlwpAZXSS3rOHOqxba9ie61qUd/DfOMay21Pbm4r6VuS7pE0SdIltjfO\n+5xk+8tK88FLaerGGttzc1yz2mxvjKvxBmpKU2z9/m/Jc8x/La9+2fZ/NFzLt0v6bt72/Yi4pcVx\nj6nOWZu4JjbFVr2lc3xTXK0eNP5Ppd7y10maldsrzzP/I6VhV49KOr7Fcac1XMsZLbav1BTXxL5N\n/eJqvhmVpIuUnk0YI+lC21vnhqvYPlTSJ6v4I2JRi2PPznHNbnXOAIxCEcHCwsLS1SJprlLv5vRB\n1P2MpCW5fijNGPEPpYdEq7I5TW2eJ+nuhu2LlHrPQ+mFRTPaxZC3haTZLWKZ3rDPAWNvantMQ9sl\nSgnbPxvKQtItkqa1ad/x2A1xD7TUvlduv5nSeOSq3mMN53iJpCMGcT1ntdk+mLii1XdXSpR/2lDn\naaUbqmr9j5ImDXTO22yfNci4jmnT/oNKiXnjNW38e7pjm3bTGurNGOBad1rmttn/epL+2lDviYY4\nQ9KpHa7l7E5/T1hYWEbfQk85gJ6IiOOUEsbTJf1ZfdMYzpN0maTDJW3X1OYfkraWdKqk6iHPpyT9\nWNIO0X+mll65QNIXlIYM3Ks0texqSt/j55I+JGnLiJjbpn01rGShRuABvUgvaNpU0teVErpVJD2i\nlBDvFBFfatUuDweqHpq9dgTi+pfSFJH7S7paaThLSLpJ0kxJ20b7WVeqc3Zdm+3Dje0sSa9Xmn7z\nIaVpCO9V+qVj84j41QBxhaTrRyCu+yRtLum/lB7kHaeUmP9a0h4RcUCH5lVsS/1aAhgZjujlEDkA\nWLHZPk3SRyWdGBGHDVS/V/LwiKsk3S9pg4h4pnBIz7F9p9KY810i4pLS8VRsHyHpi5L+OyLeUzqe\niu2pSjeJCyWtH+nFVgBGOXrKAaC3dlDq7T+hdCBNdsifx4+yhPz5Sgn5DaMpIc92UOol/1zpQJpU\n1/I0EnJg2UFSDgA9YnttSRtJ+lZEPDRQ/R7bXql39YzSgTSpZrIZVYlvflB1G0kXRYuHUwvbXqPz\nxg9ABwxfAQAAAAqjpxwAAAAojKQcAAAAKIykHAAAACiMpBwAAAAojKQcAAAAKIykHMAKx/Y022Gb\n6acAAKMCSTkADIHt6VVCP8Cy1sB7q+17Fdtvtf0Z2xfZfqBhfzsPYT+72/6V7UdsL7R9h+3jbE8a\nakwN+3yV7f1tf8f2LbYX57jOG8I+Xpbb/832M/n7nWf7NcOIa638fY/P3/mxhnO26iD3sbLtw23f\nZPtJ24/avsr2R2y7y7jG2H5j3u/5tuc0xLX/EPbzZtsX255v+2nbf7H9tfxSJQDLkXGlAwCAZdQS\nSQ8PsH2oXiHp0u7CSWyfLmm/vLpY0tNKLyw6StJetreLiAe62PX3JG02jLh2kvRjSavlosckTZX0\nXknvtr1PRJzdxa73lnTyMOJaXdKvJFU3BgsljZe0dV52sf2uiFg8xF1X++2a7aMkHZdXl0h6UtJL\nJX1C6VruGBG3DucYAEYPesoBoDv3RsTUDss/utzvo5KukPQlSe8eSkPbBygl5EskfUrSxIiYJOkN\nku5RSujO7zKuf0m6SdK3JX1U0mVDiGuqpB8qJeSXS5oWEVOUkvJzlDqIvmN7ky7iCkn3SbpQ6cbj\n00Nsf4ZSQv4PSbtImpjjnKF0Q/MOScd2EZckLZD0O6WbhvdJenCwDW2/XX0J+YmSpkTEZEmbKl2H\ntSVdZHuVLmMDMMrwRk8AKxzb0yTNkaSIGNLwBNvTJf1a0j0RMW0pxzUmhdT3D3PDuPe3RUTbXvSc\nnP1N0jqSTo6IQ5q2v1rS9ZIs6Z0RcfEQYxsbEc82rM+S9EFJP4iIPQdo+1VJB0m6V9LGEfFk434l\nXaOUGP8oIoZ6I9Ic13Sl6yNJ4yPi6Q5tXy3phry6a0T8pGn7QZK+qvTK+mkRMX8IcVnp/9glDWVz\nJb1E0gERcdoA7W9S+mXiwojYrWnbepLuULqB+EREfGOwcQEYvegpB9AV23Pz+Njptl9s+9u2783j\nXufY/ortyS3azcrtjsljqI/KY5SfyOVTmupPs/0N23/K46OfsH297Zm2J3SIb1XbR9u+M8c0L49f\n3ngkzsfSEBFLovuekjcrJeSh1LPavO8bJf0yr76/i9ieHbhWXb7RqJL2UxsT8ob9npRX35GHk4x4\nXNn78uefmhPy7HSlYTbjJe3WYnunuKIxIR+K/ItBNVTohBb7vk/S9/PqkK8lgNGJpBzAcG0o6TpJ\nH5I0RSkpnCbpUEnX2V63TbtVJf1W6Sf6jSTVkivbuyn1CB4o6eV536tI2kJpeMdVrR54sz1R0mxJ\nn5P0b0q9w6spjV/+o6TXt/sybpiZxfaMjt98dHlj/rw1Iu5vU6cacrJjD+KpbCypukbthrz8In+u\nLGnbEY+oT3XOftFqY0Q8pTT8ROrtOaviekzpV4RWqnP5uvz3HcAyjqQcwHB9RSl52C6PX54g6d8l\n/V0pYT+rTbv/q5Ro76k09nmKUjK/QJJsbynpPKXxxl+QtF5ETFDqtdxG6UbglUoPIDY7WdJWSsMO\n9sn7n6zU+3iHpFOH9Y2TtW3fYHtBXu6yfbrtVy6FfXej+gXgtg51bs+fa7uL2WG6VMUVDcfvJyL+\nLml+U/0RlYeXbJRXB3POevkLS3WsOzr0tldxNX4PAMswknIAw7WK0njnK6XnhmBcJGmPvH0n2616\nPydKem9E/CAiFuW290TEv/L2kyWtJOnAiPhM1fsbEc9GxFWS3ippnqS32H5ttVPbL5G0b179WETM\nqvYZEbfkdouWwvdeTdKrJT2jdOPwMqWHLG+0fdhS2P9QVb9IdJpZpXFbu18wlrbqOP/sNL5bfbH1\nKq7VlW4gG4/dSq/jajzWaLuWAEYQSTmA4To/Iu5uLoyIX0v6Q159T4t2t0REy2EDtjdQmjHkUUnf\naVUnz27y87y6U8Om3ZT+bXtALXrRc7u2PeURMTcinJdZLao8qjTO97VKDxI+TylB30Hp+46VdILt\n97VoO5KqBPOpDnUWNvy5V0MeBhOX1Bdbr+OSBnfOejlEZLReSwAjiHnKAQzX7A7bfqM01GSLFtuu\n6tBum/w5UdJ9Hd7fUiUjL2ooq471uw4//f+mw7E7ioiblKakayx7VtJvbb9RaW7qN0g63vZ53T7s\nBwBYsZCUAxiudg8VNm5bu8W2Ti/eqX6OH6e+hwQ7Wa3hz9WxOv303ynmrkXEIttHKyXm6ykNb7l+\nJI7VwoL8Ob5Dncbz9GTbWkvXYOKS+mLrdVzS4M5Zr+KSRu+1BDCCGL4CoJROU9lV/zbd3DCUpNMy\nowfxDlbjbBkv7eFxq5uQF3So07ht3gjG0qiKa40BXntfxdaruB5XX/I7mHPWq7ik0XstAYwgknIA\nwzWYxKFTr3grD+XPF3Ws1Vp1rMEmNMuLajaOTm/FrGb1eDjPeNILjbOEtJzBJM8Es05T/RGV54O/\nI68O5pz1JK6mY70iz/PeSuOsNne0qQNgGUJSDmC4dhjEths61GmlGm/+PNtbDbFtdaxt3X4weqeY\nh6sx3jkjeJxm1VssN+kwN/xb8ucVPYincof6brJ2alOnKl8k6coRj6hPdc5axpV79rfLq708Z1Vc\nkyVt2aZOdS2viYgFbeoAWIaQlAMYrvfarg3TsL290gOPknTBUHYYEXdKujqvftn2Su3q2h6fXzFf\n+ZGkJZJeKGnvFvXXkLT/UOJpat/2qdMc5+fy6jwN/WZkOK5Qmut7jNKLm/qxvZnSWz8l6ZxeBZUf\ndD0vr36s+S2suSf44Lx6cUQ83qvY1PdWzI1sv6PF9v2UEuOnJF3Yq6Ai4nZJN+fVTzVvt/0CSXvl\n1Z5dSwAji6QcwHAtkvRz29tIKcmyvYukH+btl0fE77vY7yeU5gDfXtIVtretfsq3Pdb2K21/VtJf\n1TBPc0TcI+nMvHqa7Q9USX1+sc+lSm8TbWkQb/S81fbHbb+sStBzPNsqJcbVnOxHtpp5xfbsvO/Z\nbY6/hu21qqVh0+qN5c03KhHxjKRj8urBtg+tblZsv14pqRwj6fcRcUmL485o+N7TWmxfrSmu6kZo\n5aa4Wk3P9yWlMdwvlvQj2y/O+1xb0iyl3uBFkv6zzTmp4jqmxbYxTXFNbti8ZptzKUmKiBslnZ9X\nZ9l+e97nWNsfkHR83nZyRMxvbm97Vo5rbpu4Jzcdv/o/d0LTOVulRfNP58932/6y7Ul5nxtLuljS\nJKW/+2e0OO70hnM2vVVsAEahiGBhYWEZ8iJprtJ41g8r9dCGpCeU5k+OvPxZ0rpN7WblbccM4hhv\nU5oXvNrf00pvCl3UUBaSXtLUbqJST3tju2o/Tyr1Moby0OKmttMa2s1osT2a9vuw0s1DVfYvSTM7\nfKfZud7sAc7rQMv0Nu1Pb6izKF+Tav0vkl7Qpt2MhnrTWmw/ZpBxzWqz/52UHqys6j2q9ItGdc72\n7nDOqja1vzNN16vj0mbfqyu9HbaqtyBf12r9Yknj2rSdlevMHeBaD7TU/p7l9p9pqLNY6c251frD\nkjZt0276QH9PWFhYRt9CTzmA4bpb6UU6ZyolDWOVEssTJb02IrqeGSIifi7p5ZKOUxoK8oykKUq9\nrn9Q6oF9TaTe8cZ2TyolJp+VdFcuflrSDyS9Tp3nSB/IR5VeSnRbjmNKjut/JJ0iabOIOL598+d6\n9a8dRgxtRcRHJL1XaVzyk0rTSt4p6QuSNo+IdlNFVnHdrxGYzSMiLpe0uaTvSrpPabq/h5R6qreO\niLNbtWsaH7/Uz1mk4TLbSDpCachIKF3Pq5Wu9TsjYnGb5iN9LY9Tupn5qaR/Kv068VdJX1dKyG8d\nIK6F6u0DqgCGwRFROgYAy6D8k/1LJL0xImaXjWbZYHuqUsK7UNL60WJIRCm2L5X0Vkkfj4hTSsdT\nsb2n0tjv6yPitaXjqdgep5QoT1C6EfufwiE9x/ZpSjcUJ0bEYaXjATA49JQDQO9Us76cNsoS8rFK\nD+U+oBZjlAurztmxRaOoe43SMKkLR1NCnu2g9HDqCaUDATB4JOUA0Dvba3QmS1soJZhfjvTA6Giy\nvaQbI+Li0oE02T5/fr5oFE3yw7MbSfpWRDw0UH0AowfDVwB0heErAAAsPfSUAwAAAIXRUw4AAAAU\nRk85AAAAUBhJOQAAAFAYSTkAAABQGEk5AAAAUBhJOQAAAFAYSTkAAABQ2LjSAYwU28z1CAAAgBEX\nER7uPugpBwAAAAojKQcAAAAKIykHAAAACiMpBwAAAAojKQcAAAAKIykHAAAACiMpBwAAAAojKQcA\nAAAKIykHAAAACiMpBwAAAAojKQcAAAAKIykHAAAACiMpBwAAAAojKQcAAAAKIykHAAAACiMpBwAA\nAAojKQcAAAAKIykHAAAACiMpBwAAAAojKQcAAAAKIykHAAAACiMpBwAAAAojKQcAAAAKG1c6ACxd\nG264Ya1sk002qZXdf//9tbJTTjml3/pWW21VqxMRg4rjkUce6be+++671+rccMMNtbLHH398UPsH\nAABYntBTDgAAABRGUg4AAAAURlIOAAAAFEZSDgAAABTmwT64t6yxvXx+sQYzZ86slR177LG1spVW\nWqlW9uyzz9bKxo4d22/9wQcfrNWZOnXqUELseLzFixfXyhYuXFgru+CCC2plp556ar/1VrHOnz9/\nKCECAAB0JSI83H3QUw4AAAAURlIOAAAAFEZSDgAAABTGy4NGqUmTJtXKdt11137rG2ywQa1Oq/Hj\nDzzwQK3s8ssvr5Vde+21/dZ/85vf1OrMmzevHuwgvOpVr6qVbbzxxrWyVVddtVa2+eab18rOOuus\nfusf//jHa3UYUw4AAJYV9JQDAAAAhZGUAwAAAIWRlAMAAACFkZQDAAAAhfHyoFFg++23r5UdcMAB\ntbI99thjwH3NmTOnVrbzzjvXyu6+++5BRjc6jR8/vt/6U089VSgSAACwouPlQQAAAMBygKQcAAAA\nKIykHAAAACiMpBwAAAAojAc9e6zVmzovvfTSWtmmm25aK5s4cWK/9UsuuaRWZ999962VPfLII0MJ\nEQAAAEPAg54AAADAcoCkHAAAACiMpBwAAAAojKQcAAAAKGxc6QBWNCeddFKtbOuttx5U29mzZ/db\nP/TQQ2t1eKgTAABg2UNPOQAAAFAYSTkAAABQGEk5AAAAUBhJOQAAAFAYb/QcQa3eynndddfVylZa\naaVa2Zw5c2plO++8c7/1u+++exjRAQAAYGngjZ4AAADAcoCkHAAAACiMpBwAAAAojJcHjaDFixfX\nygY7hv/ggw+ulTGGHAAAYPlETzkAAABQGEk5AAAAUBhJOQAAAFAYSTkAAABQGC8P6rHHH3+8VjZh\nwoRa2bRp02pl995770iEBAAAgGHg5UEAAADAcoCkHAAAACiMpBwAAAAojKQcAAAAKIwHPXts//33\nr5WdfPLJtbKrr766VnbggQf2W7/tttuWXmAAAADoCg96AgAAAMsBknIAAACgMJJyAAAAoDCScgAA\nAKAwHvQcBd72trfVyg4//PBa2Q477NBv/ZprrqnV2W+//Wplt9566zCiAwAAQCc86AkAAAAsB0jK\nAQAAgMJIygEAAIDCSMoBAACAwnjQc5QaM6Z+v/SmN72p3/pRRx1Vq7P55pvXyq677rpa2TnnnFMr\n++53vzuUEAEAACAe9AQAAACWCyTlAAAAQGEk5QAAAEBhjClfho0dO7ZWtuWWW9bKDjnkkFrZtttu\nWytbeeWV+61/8pOfrNU599xza2VLlizpGCcAAMDyjDHlAAAAwHKApBwAAAAojKQcAAAAKIykHAAA\nACiMBz1XUOuss06t7Mgjj+y3vvfee9fqnH/++bWygw46qFa2ePHiYUQHAACw7OBBTwAAAGA5QFIO\nAAAAFEZSDgAAABRGUg4AAAAUNq50AChj/vz5tbKDDz643/pdd91Vq3PKKafUytZcc81a2Z577jmM\n6AAAAFYs9JQDAAAAhZGUAwAAAIWRlAMAAACFkZQDAAAAhfFGTzxnk0026bd+5ZVX1uqsvvrqtbIF\nCxbUyjbccMN+660eLAUAAFge8EZPAAAAYDlAUg4AAAAURlIOAAAAFMbLg1YAY8bU77222WabWtlx\nxx3Xb33y5Mm1Oq2eQbj55ptrZY8++uhQQgQAAFih0VMOAAAAFEZSDgAAABRGUg4AAAAURlIOAAAA\nFMaDnj125JFH1souu+yyWtkNN9ww4L7Gjx9fK9tss81qZXvttVet7MADDxxw/4sWLaqVnX322bWy\no48+elBtAQAA0Bo95QAAAEBhJOUAAABAYSTlAAAAQGEk5QAAAEBhPOg5Cpx55pm1sgcffHDAduuv\nv36tbMMNN+w6jj//+c/91mfOnFmrc9FFF3W9fwAAALRGTzkAAABQGEk5AAAAUBhJOQAAAFAYSTkA\nAABQmCOidAwjwvYy88X22WefWtnXv/71Wtlqq63W1f6bH+CUpAsuuKBW9s1vfrPf+rx587o6HgAA\nwIokIjzcfdBTDgAAABRGUg4AAAAURlIOAAAAFEZSDgAAABTGg54AAADAMPCgJwAAALAcICkHAAAA\nCiMpBwAAAAojKQcAAAAKIykHAAAACiMpBwAAAAojKQcAAAAKIykHAAAACiMpBwAAAAojKQcAAAAK\nIykHAAAACiMpBwAAAAojKQcAAAAKc0SUjgEAAABYodFTDgAAABRGUg4AAAAURlIOAAAAFEZSDgAA\nABRGUg4AAAAURlIOAAAAFEZSDgAAABRGUg4AAAAURlIOAAAAFEZSDgAAABRGUg4AAAAURlIOAAAA\nFEZSDgAAABRGUg4AAAAURlIOAAAAFEZSDgAAABRGUg4AAAAURlIOAAAAFEZSDgAAABRGUg4AAAAU\nRlIOAAAAFEZSDgAAABRGUg4AAAAURlIOAAAAFEZSDgAAABRGUg4AAAAURlIOAAAAFEZSDgAAABRG\nUg4AAABgbybiAAAAfklEQVQURlIOAAAAFEZSDgAAABRGUg4AAAAURlIOAAAAFEZSDgAAABRGUg4A\nAAAURlIOAAAAFEZSDgAAABRGUg4AAAAURlIOAAAAFEZSDgAAABRGUg4AAAAURlIOAAAAFEZSDgAA\nABRGUg4AAAAURlIOAAAAFEZSDgAAABT2v6QPB46f6gkaAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fbbe1a0b940>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 127,
       "width": 370
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def get_result(result):\n",
    "    # 将 one_hot 编码解码\n",
    "    resultstr = ''\n",
    "    for i in range(n_len):\n",
    "        resultstr += str(np.argmax(result[i])) + ','\n",
    "    return resultstr\n",
    "\n",
    "index = random.randint(0, n_test-1)\n",
    "y_pred = model.predict(X_test[index].reshape(1, height, width*n_len, 1))\n",
    "\n",
    "plt.title('real: %s\\npred:%s'%(get_result([y_test[x][index] for x in range(n_len)]), get_result(y_pred)))\n",
    "plt.imshow(X_test[index,:,:,0], cmap='gray')\n",
    "plt.axis('off')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 保存模型\n",
    "\n",
    "模型达到满意的效果以后，我们需要将模型的权值和结构保存，以便下次调用。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "model.save_weights('model.h5')\n",
    "with open('model.json', 'w') as f:\n",
    "    f.write(model.to_json())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "reference\n",
    "* [machinelearningmastery](http://machinelearningmastery.com/handwritten-digit-recognition-using-convolutional-neural-networks-python-keras/)\n",
    "* [kaggle](https://www.kaggle.com/poonaml/digit-recognizer/deep-neural-network-keras-way)\n",
    "* [Floyd](https://www.floydhub.com/)"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}