minist_MLP_tensorflow.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "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": [
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "import tensorflow as tf\n",
    "mnist = input_data.read_data_sets(\"MNIST_data/\", one_hot=True)\n",
    "sess = tf.InteractiveSession()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "in_units = 784\n",
    "hl_units = 300\n",
    "Wl = tf.Variable(tf .truncated_normal([in_units, hl_units], stddev=0.1))\n",
    "bl = tf.Variable(tf.zeros([hl_units]))\n",
    "W2 = tf.Variable(tf.zeros([hl_units, 10]))\n",
    "b2 = tf.Variable(tf.zeros([10]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x = tf.placeholder(tf.float32, [None, in_units])\n",
    "keep_prob = tf.placeholder(tf.float32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "hiddenl = tf.nn.relu(tf.matmul(x, Wl) + bl)\n",
    "hiddenl_drop = tf.nn.dropout(hiddenl, keep_prob)\n",
    "y = tf.nn.softmax(tf.matmul(hiddenl_drop, W2) + b2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y_ = tf.placeholder (tf.float32, [None, 10])\n",
    "cross_entropy = tf.reduce_mean(-tf.reduce_sum(y_ * tf.log(y),reduction_indices=[1]))\n",
    "train_step = tf.train.AdagradOptimizer(0.3 ).minimize(cross_entropy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "tf.global_variables_initializer().run()\n",
    "for i in range(3000):\n",
    "    batch_xs, batch_ys = mnist .train.next_batch(100)\n",
    "    train_step.run({x: batch_xs , y_ : batch_ys , keep_prob: 0.75} )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9779\n"
     ]
    }
   ],
   "source": [
    "correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1))\n",
    "accuracy = tf.reduce_mean(tf.cast (correct_prediction, tf.float32))\n",
    "print(accuracy.eval({x: mnist.test.images, y_: mnist.test.labels,keep_prob : 1.0}))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [Anaconda3]",
   "language": "python",
   "name": "Python [Anaconda3]"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}