keras-mnist.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 20180110 - My Neural Network Lab notes - MNIST dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import keras\n",
    "from keras.datasets import mnist\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Activation\n",
    "from keras.preprocessing.text import Tokenizer\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.cm as cm\n",
    "%matplotlib inline\n",
    "\n",
    "np.random.seed(666)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data\n",
    "Keras include sample datasets, e.g. MNIST database of handwritten digits\n",
    "\n",
    "http://yann.lecun.com/exdb/mnist/\n",
    "\n",
    "Dataset of 60,000 28x28 grayscale images of the 10 digits, along with a test set of 10,000 images.\n",
    "\n",
    "Usage:\n",
    "\n",
    "from keras.datasets import mnist\n",
    "\n",
    "(x_train, y_train), (x_test, y_test) = mnist.load_data()\n",
    "\n",
    "Returns:\n",
    "\n",
    "2 tuples:\n",
    "\n",
    "x_train, x_test: uint8 array of grayscale image data with shape (num_samples, 28, 28).\n",
    "\n",
    "y_train, y_test: uint8 array of digit labels (integers in range 0-9) with shape (num_samples,).\n",
    "\n",
    "Arguments:\n",
    "\n",
    "path: if you do not have the index file locally (at '~/.keras/datasets/' + path), it will be downloaded to this location.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Length of data vector\n",
      "60000\n"
     ]
    }
   ],
   "source": [
    "(x_train, y_train), (x_test, y_test) = mnist.load_data()\n",
    "\n",
    "print('\\nLength of data vector')\n",
    "print(len(x_train))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABHEAAAEUCAYAAACs6aoPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAy+G7Jki3gHZAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fefa321f748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(20,20))\n",
    "for i in range(4):\n",
    "    ax = fig.add_subplot(1, 4, i+1, xticks=[], yticks=[])\n",
    "    ax.imshow(x_train[i], cmap='gray')\n",
    "    ax.set_title(str(y_train[i]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data normalization\n",
    "\n",
    "Converting [0,255] --> [0,1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x_train = x_train.astype('float32') / 255\n",
    "x_test = x_test.astype('float32') / 255 "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## One-hot encoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.  0.  0.  0.  0.  1.  0.  0.  0.  0.]\n",
      " [ 1.  0.  0.  0.  0.  0.  0.  0.  0.  0.]\n",
      " [ 0.  0.  0.  0.  1.  0.  0.  0.  0.  0.]\n",
      " [ 0.  1.  0.  0.  0.  0.  0.  0.  0.  0.]]\n"
     ]
    }
   ],
   "source": [
    "from keras.utils import np_utils\n",
    "\n",
    "y_train = np_utils.to_categorical(y_train, 10)\n",
    "y_test = np_utils.to_categorical(y_test, 10)\n",
    "\n",
    "print(y_train[:4])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(28, 28)\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "flatten_1 (Flatten)          (None, 784)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 32)                25120     \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 32)                1056      \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 32)                0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 10)                330       \n",
      "=================================================================\n",
      "Total params: 26,506.0\n",
      "Trainable params: 26,506.0\n",
      "Non-trainable params: 0.0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Flatten\n",
    "\n",
    "input_shape=x_train.shape[1:]\n",
    "print(input_shape)\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Flatten(input_shape=(28,28)))\n",
    "model.add(Dense(32, activation='relu'))\n",
    "model.add(Dense(32, activation='relu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(10, activation='softmax'))\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compile & train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "60000/60000 [==============================] - 2s - loss: 0.7416 - acc: 0.7648     \n",
      "Epoch 2/10\n",
      "60000/60000 [==============================] - 2s - loss: 0.4170 - acc: 0.8773     \n",
      "Epoch 3/10\n",
      "60000/60000 [==============================] - 2s - loss: 0.3486 - acc: 0.8979     \n",
      "Epoch 4/10\n",
      "60000/60000 [==============================] - 2s - loss: 0.3112 - acc: 0.9095     \n",
      "Epoch 5/10\n",
      "60000/60000 [==============================] - 2s - loss: 0.2813 - acc: 0.9197     \n",
      "Epoch 6/10\n",
      "60000/60000 [==============================] - 2s - loss: 0.2638 - acc: 0.9254     \n",
      "Epoch 7/10\n",
      "60000/60000 [==============================] - 2s - loss: 0.2527 - acc: 0.9294     \n",
      "Epoch 8/10\n",
      "60000/60000 [==============================] - 2s - loss: 0.2453 - acc: 0.9322     \n",
      "Epoch 9/10\n",
      "60000/60000 [==============================] - 2s - loss: 0.2343 - acc: 0.9358     \n",
      "Epoch 10/10\n",
      "60000/60000 [==============================] - 2s - loss: 0.2313 - acc: 0.9359     \n"
     ]
    }
   ],
   "source": [
    "model.compile(loss='categorical_crossentropy', optimizer='rmsprop', metrics=['accuracy'])\n",
    "\n",
    "hist = model.fit(x_train, y_train, batch_size=64, epochs=10, verbose=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Accuracy of the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 95.9100%\n"
     ]
    }
   ],
   "source": [
    "score = model.evaluate(x_test, y_test, verbose=0)\n",
    "accuracy = 100*score[1]\n",
    "print('Accuracy: %.4f%%' % accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}