23_tf_mnist_u1000_u1000.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.simplefilter(action='ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from time import time\n",
    "\n",
    "import tensorflow as tf\n",
    "tf.logging.set_verbosity(tf.logging.ERROR)  # 过滤掉 Tensorflow 的 Warning 信息\n",
    "\n",
    "import tensorflow.examples.tutorials.mnist.input_data as input_data\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. 数据预处理"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting data/train-images-idx3-ubyte.gz\n",
      "Extracting data/train-labels-idx1-ubyte.gz\n",
      "Extracting data/t10k-images-idx3-ubyte.gz\n",
      "Extracting data/t10k-labels-idx1-ubyte.gz\n"
     ]
    }
   ],
   "source": [
    "mnist = input_data.read_data_sets('data/', one_hot=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train images shape: (55000, 784)\n",
      "train labels shape: (55000, 10)\n",
      "\n",
      "validation images shape: (5000, 784)\n",
      "validation labels shape: (5000, 10)\n",
      "\n",
      "test images shape: (10000, 784)\n",
      "test labels shape: (10000, 10)\n"
     ]
    }
   ],
   "source": [
    "print('train images shape:', mnist.train.images.shape)\n",
    "print('train labels shape:', mnist.train.labels.shape)\n",
    "print()\n",
    "print('validation images shape:', mnist.validation.images.shape)\n",
    "print('validation labels shape:', mnist.validation.labels.shape)\n",
    "print()\n",
    "print('test images shape:', mnist.test.images.shape)\n",
    "print('test labels shape:', mnist.test.labels.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. 建立模型"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2.1 定义 layer 函数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def layer(output_dim, input_dim, inputs, activation=None):\n",
    "    W = tf.Variable(tf.random_normal([input_dim, output_dim]))\n",
    "    b = tf.Variable(tf.random_normal([1, output_dim]))\n",
    "    result = tf.matmul(inputs, W) + b\n",
    "    \n",
    "    if activation is None:\n",
    "        outputs = result\n",
    "    else:\n",
    "        outputs = activation(result)\n",
    "    \n",
    "    return outputs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2.2 建立输入层"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = tf.placeholder('float', [None, 784])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2.3 建立隐藏层"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "h1 = layer(output_dim=1000, input_dim=784, inputs=x, activation=tf.nn.relu)  # 隐藏层的神经元增大为1000\n",
    "h2 = layer(output_dim=1000, input_dim=1000, inputs=h1, activation=tf.nn.relu)  # 增加隐藏层2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2.4 建立输出层"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_predict = layer(output_dim=10, input_dim=1000, inputs=h2, activation=None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. 定义训练方式"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.1 建立训练数据 label 真实值的 placeholder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_label = tf.placeholder('float', [None, 10])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.2 定义损失函数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "loss_function = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=y_predict, labels=y_label))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.3 定义优化器"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "optimizer = tf.train.AdamOptimizer(learning_rate=0.001).minimize(loss_function)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4. 定义评估模型准确率的方式"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4.1 计算每一项数据是否预测正确"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "correct_prediction = tf.equal(tf.argmax(y_label, axis=1), tf.argmax(y_predict, axis=1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4.2 计算预测正确结果的平均值"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "accuracy = tf.reduce_mean(tf.cast(correct_prediction, 'float'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5. 开始训练"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_epochs = 15\n",
    "batch_size = 100\n",
    "total_batch = int(mnist.train.num_examples / batch_size)\n",
    "epoch_list = []\n",
    "loss_list = []\n",
    "acc_list = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train epoch: 01 loss: 135.302246094 acc: 0.9192\n",
      "train epoch: 02 loss: 87.908386230 acc: 0.9372\n",
      "train epoch: 03 loss: 85.802574158 acc: 0.9426\n",
      "train epoch: 04 loss: 68.124221802 acc: 0.9496\n",
      "train epoch: 05 loss: 62.317188263 acc: 0.9526\n",
      "train epoch: 06 loss: 67.303459167 acc: 0.9514\n",
      "train epoch: 07 loss: 60.414093018 acc: 0.9604\n",
      "train epoch: 08 loss: 53.442119598 acc: 0.9602\n",
      "train epoch: 09 loss: 58.059757233 acc: 0.9576\n",
      "train epoch: 10 loss: 66.963310242 acc: 0.9522\n",
      "train epoch: 11 loss: 59.823623657 acc: 0.9584\n",
      "train epoch: 12 loss: 53.104755402 acc: 0.9662\n",
      "train epoch: 13 loss: 48.553176880 acc: 0.9676\n",
      "train epoch: 14 loss: 56.440586090 acc: 0.9646\n",
      "train epoch: 15 loss: 49.556381226 acc: 0.9692\n",
      "\n",
      "train finished. takes 182.19251680374146 seconds\n"
     ]
    }
   ],
   "source": [
    "start_time = time()\n",
    "sess = tf.Session()\n",
    "sess.run(tf.global_variables_initializer())\n",
    "\n",
    "for epoch in range(train_epochs):\n",
    "    for i in range(total_batch):\n",
    "        x_batch, y_batch = mnist.train.next_batch(batch_size=batch_size)\n",
    "        sess.run(optimizer, feed_dict={x: x_batch, y_label: y_batch})\n",
    "    \n",
    "    loss, acc = sess.run([loss_function, accuracy], \n",
    "                         feed_dict={x: mnist.validation.images, y_label: mnist.validation.labels})\n",
    "    epoch_list.append(epoch+1)\n",
    "    loss_list.append(loss)\n",
    "    acc_list.append(acc)\n",
    "    print('train epoch:', '%02d' % (epoch + 1), 'loss:', '{:.9f}'.format(loss), 'acc:', acc)\n",
    "\n",
    "print()\n",
    "print('train finished. takes', time() - start_time, 'seconds')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6. 以图形显示训练过程"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def show_train_history(x_values, y_values, title):\n",
    "    plt.plot(x_values, y_values, label=title)\n",
    "    plt.xlabel('Epoch')\n",
    "    plt.ylabel(title)\n",
    "    plt.legend([title], loc='upper left')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "show_train_history(epoch_list, acc_list, 'acc')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "show_train_history(epoch_list, loss_list, 'loss')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 7. 评估模型准确率"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy: 0.9654\n"
     ]
    }
   ],
   "source": [
    "print('accuracy:', sess.run(accuracy, feed_dict={x: mnist.test.images, y_label: mnist.test.labels}))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 8. 进行预测"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 8.1 执行预测"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "prediction_result = sess.run(tf.argmax(y_predict, axis=1), feed_dict={x: mnist.test.images})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 8.2 预测结果"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([7, 2, 1, 0, 4, 1, 4, 9, 5, 9])"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prediction_result[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 8.3 定义函数以显示10项预测结"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_images_labels_prediction(images, labels, predictions, idx, num=10):\n",
    "    \"\"\"\n",
    "    images: 数字图像数组\n",
    "    labels: 真实值数组\n",
    "    predictions: 预测结果数据\n",
    "    idx: 开始显示的数据index\n",
    "    num: 要显示的数据项数, 默认为10, 不超过25\n",
    "    \"\"\"\n",
    "    fig = plt.gcf()\n",
    "    fig.set_size_inches(12, 14)\n",
    "    if num > 25:\n",
    "        num = 25\n",
    "    for i in range(0, num):\n",
    "        ax = plt.subplot(5, 5, i+1)\n",
    "        ax.imshow(images[idx].reshape(28, 28), cmap='binary')\n",
    "        title = 'lable=' + str(np.argmax(labels[idx]))\n",
    "        if len(predictions) > 0:\n",
    "            title += ',predict=' + str(predictions[idx])\n",
    "        ax.set_title(title, fontsize=10)\n",
    "        ax.set_xticks([])\n",
    "        ax.set_yticks([])\n",
    "        idx += 1\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x1008 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_images_labels_prediction(mnist.test.images, mnist.test.labels, prediction_result, 0, 10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 9. 找出预测错误"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "i=18 label=3 predict=8\n",
      "i=63 label=3 predict=2\n",
      "i=92 label=9 predict=4\n",
      "i=151 label=9 predict=8\n",
      "i=193 label=9 predict=7\n",
      "i=241 label=9 predict=8\n",
      "i=245 label=3 predict=7\n",
      "i=247 label=4 predict=2\n",
      "i=259 label=6 predict=0\n",
      "i=268 label=8 predict=3\n"
     ]
    }
   ],
   "source": [
    "for i in range(300):\n",
    "    label = np.argmax(mnist.test.labels[i])\n",
    "    predict = prediction_result[i]\n",
    "    if predict != label:\n",
    "        print('i=' + str(i), 'label=' + str(label), 'predict=' + str(predict))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "sess.close()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "tensorflow-keras-practice",
   "language": "python",
   "name": "tensorflow-keras-practice"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}