keras/keras_basic_2dcnn_digit_sklearn.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CNN for Digit Recognition"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_digits\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "\n",
    "digit = load_digits()\n",
    "\n",
    "data_x = digit.data\n",
    "data_y = digit.target\n",
    "\n",
    "\n",
    "x_train, x_test, y_train, y_test = train_test_split(data_x, data_y, test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. data preprocess"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x_train shape: (1437, 8, 8, 1)\n",
      "1437 train samples\n",
      "360 test samples\n"
     ]
    }
   ],
   "source": [
    "import keras\n",
    "from keras.datasets import mnist\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Flatten\n",
    "from keras.layers import Conv2D, MaxPooling2D\n",
    "from keras import backend as K\n",
    "\n",
    "batch_size = 128\n",
    "num_classes = 10\n",
    "epochs = 12\n",
    "\n",
    "# input image dimensions\n",
    "img_rows, img_cols = 8, 8\n",
    "\n",
    "\n",
    "if K.image_data_format() == 'channels_first':\n",
    "    x_train = x_train.reshape(x_train.shape[0], 1, img_rows, img_cols)\n",
    "    x_test = x_test.reshape(x_test.shape[0], 1, img_rows, img_cols)\n",
    "    input_shape = (1, img_rows, img_cols)\n",
    "else:\n",
    "    x_train = x_train.reshape(x_train.shape[0], img_rows, img_cols, 1)\n",
    "    x_test = x_test.reshape(x_test.shape[0], img_rows, img_cols, 1)\n",
    "    input_shape = (img_rows, img_cols, 1)\n",
    "\n",
    "x_train = x_train.astype('float32')\n",
    "x_test = x_test.astype('float32')\n",
    "x_train /= 255\n",
    "x_test /= 255\n",
    "print('x_train shape:', x_train.shape)\n",
    "print(x_train.shape[0], 'train samples')\n",
    "print(x_test.shape[0], 'test samples')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. CNN model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# convert class vectors to binary class matrices\n",
    "y_train = keras.utils.to_categorical(y_train, num_classes)\n",
    "y_test = keras.utils.to_categorical(y_test, num_classes)\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Conv2D(32, kernel_size=(3, 3),\n",
    "                 activation='relu',\n",
    "                 input_shape=input_shape))\n",
    "model.add(Conv2D(64, (3, 3), activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "model.add(Dropout(0.25))\n",
    "model.add(Flatten())\n",
    "model.add(Dense(128, activation='relu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(num_classes, activation='softmax'))\n",
    "\n",
    "model.compile(loss=keras.losses.categorical_crossentropy,\n",
    "              optimizer=keras.optimizers.Adadelta(),\n",
    "              metrics=['accuracy'])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. train & test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 1437 samples, validate on 360 samples\n",
      "Epoch 1/12\n",
      "1437/1437 [==============================] - 0s - loss: 2.3006 - acc: 0.1162 - val_loss: 2.2990 - val_acc: 0.1833\n",
      "Epoch 2/12\n",
      "1437/1437 [==============================] - 0s - loss: 2.2980 - acc: 0.1524 - val_loss: 2.2963 - val_acc: 0.1528\n",
      "Epoch 3/12\n",
      "1437/1437 [==============================] - 0s - loss: 2.2950 - acc: 0.1684 - val_loss: 2.2921 - val_acc: 0.2833\n",
      "Epoch 4/12\n",
      "1437/1437 [==============================] - 0s - loss: 2.2893 - acc: 0.1921 - val_loss: 2.2851 - val_acc: 0.1611\n",
      "Epoch 5/12\n",
      "1437/1437 [==============================] - 0s - loss: 2.2814 - acc: 0.1942 - val_loss: 2.2763 - val_acc: 0.1639\n",
      "Epoch 6/12\n",
      "1437/1437 [==============================] - 0s - loss: 2.2696 - acc: 0.2053 - val_loss: 2.2617 - val_acc: 0.1972\n",
      "Epoch 7/12\n",
      "1437/1437 [==============================] - 0s - loss: 2.2478 - acc: 0.2714 - val_loss: 2.2350 - val_acc: 0.3139\n",
      "Epoch 8/12\n",
      "1437/1437 [==============================] - 0s - loss: 2.2116 - acc: 0.2853 - val_loss: 2.1809 - val_acc: 0.3722\n",
      "Epoch 9/12\n",
      "1437/1437 [==============================] - 0s - loss: 2.1375 - acc: 0.3243 - val_loss: 2.0770 - val_acc: 0.3194\n",
      "Epoch 10/12\n",
      "1437/1437 [==============================] - 0s - loss: 1.9909 - acc: 0.3974 - val_loss: 1.8549 - val_acc: 0.5222\n",
      "Epoch 11/12\n",
      "1437/1437 [==============================] - 0s - loss: 1.7795 - acc: 0.4356 - val_loss: 1.5192 - val_acc: 0.7083\n",
      "Epoch 12/12\n",
      "1437/1437 [==============================] - 0s - loss: 1.5215 - acc: 0.5031 - val_loss: 1.2430 - val_acc: 0.7222\n",
      "Test loss: 1.24302191469\n",
      "Test accuracy: 0.722222222222\n"
     ]
    }
   ],
   "source": [
    "model.fit(x_train, y_train,\n",
    "          batch_size=batch_size,\n",
    "          epochs=epochs,\n",
    "          verbose=1,\n",
    "          validation_data=(x_test, y_test))\n",
    "score = model.evaluate(x_test, y_test, verbose=0)\n",
    "print('Test loss:', score[0])\n",
    "print('Test accuracy:', score[1])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "# Convert one-hot to index\n",
    "y_pred = np.argmax(model.predict(x_train), axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. model visualize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<svg height=\"629pt\" viewBox=\"0.00 0.00 262.00 629.00\" width=\"262pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g class=\"graph\" id=\"graph0\" transform=\"scale(1 1) rotate(0) translate(4 625)\">\n",
       "<title>G</title>\n",
       "<polygon fill=\"white\" points=\"-4,4 -4,-625 258,-625 258,4 -4,4\" stroke=\"none\"/>\n",
       "<!-- 139768124402320 -->\n",
       "<g class=\"node\" id=\"node1\"><title>139768124402320</title>\n",
       "<polygon fill=\"none\" points=\"17.5,-584.5 17.5,-620.5 236.5,-620.5 236.5,-584.5 17.5,-584.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"127\" y=\"-598.8\">conv2d_1_input: InputLayer</text>\n",
       "</g>\n",
       "<!-- 139768987039112 -->\n",
       "<g class=\"node\" id=\"node2\"><title>139768987039112</title>\n",
       "<polygon fill=\"none\" points=\"51.5,-511.5 51.5,-547.5 202.5,-547.5 202.5,-511.5 51.5,-511.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"127\" y=\"-525.8\">conv2d_1: Conv2D</text>\n",
       "</g>\n",
       "<!-- 139768124402320&#45;&gt;139768987039112 -->\n",
       "<g class=\"edge\" id=\"edge1\"><title>139768124402320-&gt;139768987039112</title>\n",
       "<path d=\"M127,-584.313C127,-576.289 127,-566.547 127,-557.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"130.5,-557.529 127,-547.529 123.5,-557.529 130.5,-557.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 139768124402432 -->\n",
       "<g class=\"node\" id=\"node3\"><title>139768124402432</title>\n",
       "<polygon fill=\"none\" points=\"51.5,-438.5 51.5,-474.5 202.5,-474.5 202.5,-438.5 51.5,-438.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"127\" y=\"-452.8\">conv2d_2: Conv2D</text>\n",
       "</g>\n",
       "<!-- 139768987039112&#45;&gt;139768124402432 -->\n",
       "<g class=\"edge\" id=\"edge2\"><title>139768987039112-&gt;139768124402432</title>\n",
       "<path d=\"M127,-511.313C127,-503.289 127,-493.547 127,-484.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"130.5,-484.529 127,-474.529 123.5,-484.529 130.5,-484.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 139768124403160 -->\n",
       "<g class=\"node\" id=\"node4\"><title>139768124403160</title>\n",
       "<polygon fill=\"none\" points=\"0,-365.5 0,-401.5 254,-401.5 254,-365.5 0,-365.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"127\" y=\"-379.8\">max_pooling2d_1: MaxPooling2D</text>\n",
       "</g>\n",
       "<!-- 139768124402432&#45;&gt;139768124403160 -->\n",
       "<g class=\"edge\" id=\"edge3\"><title>139768124402432-&gt;139768124403160</title>\n",
       "<path d=\"M127,-438.313C127,-430.289 127,-420.547 127,-411.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"130.5,-411.529 127,-401.529 123.5,-411.529 130.5,-411.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 139767982283352 -->\n",
       "<g class=\"node\" id=\"node5\"><title>139767982283352</title>\n",
       "<polygon fill=\"none\" points=\"48.5,-292.5 48.5,-328.5 205.5,-328.5 205.5,-292.5 48.5,-292.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"127\" y=\"-306.8\">dropout_1: Dropout</text>\n",
       "</g>\n",
       "<!-- 139768124403160&#45;&gt;139767982283352 -->\n",
       "<g class=\"edge\" id=\"edge4\"><title>139768124403160-&gt;139767982283352</title>\n",
       "<path d=\"M127,-365.313C127,-357.289 127,-347.547 127,-338.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"130.5,-338.529 127,-328.529 123.5,-338.529 130.5,-338.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 139767981836944 -->\n",
       "<g class=\"node\" id=\"node6\"><title>139767981836944</title>\n",
       "<polygon fill=\"none\" points=\"57,-219.5 57,-255.5 197,-255.5 197,-219.5 57,-219.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"127\" y=\"-233.8\">flatten_1: Flatten</text>\n",
       "</g>\n",
       "<!-- 139767982283352&#45;&gt;139767981836944 -->\n",
       "<g class=\"edge\" id=\"edge5\"><title>139767982283352-&gt;139767981836944</title>\n",
       "<path d=\"M127,-292.313C127,-284.289 127,-274.547 127,-265.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"130.5,-265.529 127,-255.529 123.5,-265.529 130.5,-265.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 139767981835320 -->\n",
       "<g class=\"node\" id=\"node7\"><title>139767981835320</title>\n",
       "<polygon fill=\"none\" points=\"63,-146.5 63,-182.5 191,-182.5 191,-146.5 63,-146.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"127\" y=\"-160.8\">dense_1: Dense</text>\n",
       "</g>\n",
       "<!-- 139767981836944&#45;&gt;139767981835320 -->\n",
       "<g class=\"edge\" id=\"edge6\"><title>139767981836944-&gt;139767981835320</title>\n",
       "<path d=\"M127,-219.313C127,-211.289 127,-201.547 127,-192.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"130.5,-192.529 127,-182.529 123.5,-192.529 130.5,-192.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 139767981889464 -->\n",
       "<g class=\"node\" id=\"node8\"><title>139767981889464</title>\n",
       "<polygon fill=\"none\" points=\"48.5,-73.5 48.5,-109.5 205.5,-109.5 205.5,-73.5 48.5,-73.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"127\" y=\"-87.8\">dropout_2: Dropout</text>\n",
       "</g>\n",
       "<!-- 139767981835320&#45;&gt;139767981889464 -->\n",
       "<g class=\"edge\" id=\"edge7\"><title>139767981835320-&gt;139767981889464</title>\n",
       "<path d=\"M127,-146.313C127,-138.289 127,-128.547 127,-119.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"130.5,-119.529 127,-109.529 123.5,-119.529 130.5,-119.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 139767982077432 -->\n",
       "<g class=\"node\" id=\"node9\"><title>139767982077432</title>\n",
       "<polygon fill=\"none\" points=\"63,-0.5 63,-36.5 191,-36.5 191,-0.5 63,-0.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"127\" y=\"-14.8\">dense_2: Dense</text>\n",
       "</g>\n",
       "<!-- 139767981889464&#45;&gt;139767982077432 -->\n",
       "<g class=\"edge\" id=\"edge8\"><title>139767981889464-&gt;139767982077432</title>\n",
       "<path d=\"M127,-73.3129C127,-65.2895 127,-55.5475 127,-46.5691\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"130.5,-46.5288 127,-36.5288 123.5,-46.5289 130.5,-46.5288\" stroke=\"black\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import SVG\n",
    "from keras.utils.vis_utils import model_to_dot\n",
    "\n",
    "SVG(model_to_dot(model).create(prog='dot', format='svg'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}