tensorflow/tf_basic_mnist_softmax.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Get MNIST data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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": [
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "mnist = input_data.read_data_sets(\"MNIST_data/\", one_hot=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Check dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(55000, 784)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mnist.train.images.shape# training dataset, 28*28 image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(55000, 10)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mnist.train.labels.shape# training label, one-hot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((5000, 784), (10000, 784))"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mnist.validation.images.shape, mnist.test.images.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.,  0.,  0.])"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mnist.train.labels[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4bce3d3470>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.cm as cm\n",
    "from numpy import *\n",
    "\n",
    "img = mnist.train.images[0]\n",
    "#list to array\n",
    "array_img = array(img)\n",
    "#array to matrix\n",
    "m_img = array_img.reshape(28,28)\n",
    "\n",
    "\n",
    "plt.imshow(m_img, cmap=cm.binary)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SoftMax Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9185\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "\n",
    "# Model\n",
    "x = tf.placeholder(tf.float32, [None, 784]) # None means that a dimension can be of any length.\n",
    "W = tf.Variable(tf.zeros([784, 10])) # multiply the 784-dimensional image vectors by it to produce 10-dimensional vectors of evidence for the difference classes\n",
    "b = tf.Variable(tf.zeros([10]))\n",
    "y = tf.nn.softmax(tf.matmul(x, W) + b)\n",
    "\n",
    "\n",
    "# Train: cost function optimization\n",
    "y_ = tf.placeholder(tf.float32, [None, 10]) # y_true\n",
    "cross_entropy = tf.reduce_mean(-tf.reduce_sum(y_ * tf.log(y), reduction_indices=[1])) # define cost-function\n",
    "train_step = tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy)# init learn_rate=0.5\n",
    "sess = tf.InteractiveSession()\n",
    "tf.global_variables_initializer().run()\n",
    "for _ in range(1000):# training times\n",
    "    batch_xs, batch_ys = mnist.train.next_batch(100)# batch size\n",
    "    sess.run(train_step, feed_dict={x: batch_xs, y_: batch_ys})# y_: y_true\n",
    "    \n",
    "    \n",
    "# evaluation\n",
    "correct_prediction = tf.equal(tf.argmax(y,1), tf.argmax(y_,1))# y: y_pred, y_:y_true\n",
    "accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))\n",
    "print(sess.run(accuracy, feed_dict={x: mnist.test.images, y_: mnist.test.labels}))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(100, 784)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "batch_xs.shape"
   ]
  }
 ],
 "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": "3.5.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}