TF-NN.ipynb.txt

{
 "cells": [
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "TENSORLOW EXAMPLE: NEURAL NETWORK"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting MNIST_data\\train-images-idx3-ubyte.gz\n",
      "Extracting MNIST_data\\train-labels-idx1-ubyte.gz\n",
      "Extracting MNIST_data\\t10k-images-idx3-ubyte.gz\n",
      "Extracting MNIST_data\\t10k-labels-idx1-ubyte.gz\n"
     ]
    }
   ],
   "source": [
    "# Read in data\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "mnist = input_data.read_data_sets('MNIST_data', one_hot=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def TRAIN_SIZE(num):\n",
    "    print ('Total Training Images in Dataset = ' + str(mnist.train.images.shape))\n",
    "    print ('--------------------------------------------------')\n",
    "    x_train = mnist.train.images[:num,:]\n",
    "    print ('x_train Examples Loaded = ' + str(x_train.shape))\n",
    "    y_train = mnist.train.labels[:num,:]\n",
    "    print ('y_train Examples Loaded = ' + str(y_train.shape))\n",
    "    print('')\n",
    "    return x_train, y_train\n",
    "\n",
    "def TEST_SIZE(num):\n",
    "    print ('Total Test Examples in Dataset = ' + str(mnist.test.images.shape))\n",
    "    print ('--------------------------------------------------')\n",
    "    x_test = mnist.test.images[:num,:]\n",
    "    print ('x_test Examples Loaded = ' + str(x_test.shape))\n",
    "    y_test = mnist.test.labels[:num,:]\n",
    "    print ('y_test Examples Loaded = ' + str(y_test.shape))\n",
    "    return x_test, y_test\n",
    "\n",
    "def display_train_digit(num):\n",
    "    print(Y_train[num])\n",
    "    label = Y_train[num].argmax(axis=0)\n",
    "    image = X_train[num].reshape([28,28])\n",
    "    plt.title('TRAINING Example: %d  Label: %d' % (num, label))\n",
    "    plt.imshow(image, cmap=plt.get_cmap('gray_r'))\n",
    "    plt.show()\n",
    "\n",
    "def display_test_digit(num):\n",
    "    print(Y_test[num])\n",
    "    label = Y_test[num].argmax(axis=0)\n",
    "    image = X_test[num].reshape([28,28])\n",
    "    plt.title('TESTING Example: %d  Label: %d' % (num, label))\n",
    "    plt.imshow(image, cmap=plt.get_cmap('gray_r'))\n",
    "    plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Training Images in Dataset = (55000, 784)\n",
      "--------------------------------------------------\n",
      "x_train Examples Loaded = (5500, 784)\n",
      "y_train Examples Loaded = (5500, 10)\n",
      "\n",
      "Total Test Examples in Dataset = (10000, 784)\n",
      "--------------------------------------------------\n",
      "x_test Examples Loaded = (1000, 784)\n",
      "y_test Examples Loaded = (1000, 10)\n"
     ]
    }
   ],
   "source": [
    "# Define parameters for the model\n",
    "X_train, Y_train = TRAIN_SIZE(5500)\n",
    "X_test, Y_test = TEST_SIZE(1000)"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "Tensorflow functions used:\n",
    "\n",
    "placeholder       for input data\n",
    "get_variable      gets an existing variable with these parameters or create a new one\n",
    "matmul            matrix multiplication\n",
    "nn.sigmoid        sigmoid function\n",
    "nn.softmax        softmax function\n",
    "log(x)            computes natural logarithm of x element-wise \n",
    "reduce_sum        computes the sum of elements across dimensions of a tensor\n",
    "argmax            returns the index with the largest value across axes of a tensor\n",
    "equal(x,y)        returns the truth value of (x == y) element-wise\n",
    "cast              casts a tensor to a new type\n",
    "reduce_mean       computes the mean of elements across dimensions of a tensor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# === THE MODEL ===\n",
    "\n",
    "# n features, k classes\n",
    "in_dim = 784\n",
    "hid_dim= 25\n",
    "out_dim = 10\n",
    "\n",
    "# Create placeholders for features and labels\n",
    "X = tf.placeholder(tf.float32, [None, in_dim])\n",
    "y = tf.placeholder(tf.float32, [None, out_dim])\n",
    "\n",
    "# Layer 1\n",
    "W1 = tf.get_variable('W1', [in_dim,hid_dim], initializer=tf.random_normal_initializer())\n",
    "b1 = tf.get_variable('b1',[1,hid_dim], initializer=tf.random_normal_initializer())\n",
    "h1 = tf.nn.sigmoid(tf.matmul(X, W1) + b1)\n",
    "\n",
    "# Layer 2\n",
    "W2 = tf.get_variable('W2', [hid_dim,out_dim], initializer=tf.random_normal_initializer())\n",
    "b2 = tf.get_variable('b2',[1,out_dim], initializer=tf.random_normal_initializer())\n",
    "h2 = tf.nn.softmax(tf.matmul(h1, W2) + b2)\n",
    "\n",
    "# output\n",
    "h = h2\n",
    "\n",
    "# For training: loss and trainer\n",
    "loss = tf.reduce_mean(-tf.reduce_sum(y * tf.log(h), reduction_indices=[1]))\n",
    "train_step = tf.train.GradientDescentOptimizer(0.01).minimize(loss)\n",
    "\n",
    "# For testing: accuracy\n",
    "prediction = tf.argmax(h,1)\n",
    "correct_prediction = tf.equal(prediction, tf.argmax(y,1))\n",
    "accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "After training step :  1000\n",
      "Accuracy            :  0.224\n",
      "After training step :  2000\n",
      "Accuracy            :  0.355\n",
      "After training step :  3000\n",
      "Accuracy            :  0.441\n",
      "After training step :  4000\n",
      "Accuracy            :  0.516\n",
      "After training step :  5000\n",
      "Accuracy            :  0.565\n",
      "After training step :  6000\n",
      "Accuracy            :  0.593\n",
      "After training step :  7000\n",
      "Accuracy            :  0.619\n",
      "After training step :  8000\n",
      "Accuracy            :  0.63\n",
      "After training step :  9000\n",
      "Accuracy            :  0.64\n",
      "After training step :  10000\n",
      "Accuracy            :  0.65\n"
     ]
    }
   ],
   "source": [
    "# Launch the graph\n",
    "sess = tf.Session()\n",
    "sess.run(tf.global_variables_initializer())\n",
    "    \n",
    "for i in range(10000):\n",
    "    sess.run(train_step, feed_dict={X: X_train, y: Y_train})\n",
    "    if ((i+1)%1000 == 0):\n",
    "        print('After training step : ', i+1)\n",
    "        print('Accuracy            : ', sess.run(accuracy, feed_dict={X: X_test, y: Y_test}))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.  0.  0.  0.  0.  0.  0.  1.  0.  0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nDr9Zov4fAhM80p9eyzcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xbffea90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  7\n",
      "=================\n",
      "[ 0.  0.  1.  0.  0.  0.  0.  0.  0.  0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nWaL+HxyGTSfDF74pAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xc1df550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  2\n",
      "=================\n",
      "[ 0.  1.  0.  0.  0.  0.  0.  0.  0.  0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n7PqGI7hiAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xc7a1780>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  1\n",
      "=================\n",
      "[ 1.  0.  0.  0.  0.  0.  0.  0.  0.  0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xcb07390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  0\n",
      "=================\n",
      "[ 0.  0.  0.  0.  1.  0.  0.  0.  0.  0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n4TdLlMNvliiH3yxRDr9Zov4fouJHlqKEjWgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xcea6710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  4\n",
      "=================\n",
      "[ 0.  1.  0.  0.  0.  0.  0.  0.  0.  0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xd131cc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  1\n",
      "=================\n",
      "[ 0.  0.  0.  0.  1.  0.  0.  0.  0.  0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xd39b6a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  4\n",
      "=================\n",
      "[ 0.  0.  0.  0.  0.  0.  0.  0.  0.  1.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZoly+M0S5fCbJer/AFZ5jWL8EUDTAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xd60cac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  9\n",
      "=================\n",
      "[ 0.  0.  0.  0.  0.  1.  0.  0.  0.  0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nOfxmiXL4zRLl8Jsl6v8BWd6QFGGdTqYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xd675550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  2\n",
      "=================\n",
      "[ 0.  0.  0.  0.  0.  0.  0.  0.  0.  1.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xe8c3c88>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  9\n",
      "=================\n"
     ]
    }
   ],
   "source": [
    "# Show some predictions\n",
    "Prediction = sess.run(prediction, feed_dict={X: X_test, y: Y_test})\n",
    "for i in range(10):\n",
    "    display_test_digit(i)\n",
    "    print('Prediction: ', Prediction[i])\n",
    "    print('=================')\n",
    "sess.close()"
   ]
  }
 ],
 "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}