3.4 Integrated Gradients.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3.4 Integrated Gradients"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tensorflow Walkthrough"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. Import Dependencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "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": [
    "import os\n",
    "\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "from tensorflow.python.ops import nn_ops, gen_nn_ops\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "\n",
    "from models.models_3_4 import MNIST_CNN\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "mnist = input_data.read_data_sets(\"MNIST_data/\", one_hot=True)\n",
    "\n",
    "images = mnist.train.images\n",
    "labels = mnist.train.labels\n",
    "\n",
    "logdir = './tf_logs/3_4_IG/'\n",
    "ckptdir = logdir + 'model'\n",
    "\n",
    "if not os.path.exists(logdir):\n",
    "    os.mkdir(logdir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. Building Graph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "with tf.name_scope('Classifier'):\n",
    "\n",
    "    # Initialize neural network\n",
    "    DNN = MNIST_CNN('CNN')\n",
    "\n",
    "    # Setup training process\n",
    "    X = tf.placeholder(tf.float32, [None, 784], name='X')\n",
    "    Y = tf.placeholder(tf.float32, [None, 10], name='Y')\n",
    "\n",
    "    activations, logits = DNN(X)\n",
    "    \n",
    "    tf.add_to_collection('IG', X)\n",
    "    tf.add_to_collection('IG', logits)\n",
    "    \n",
    "    for activation in activations:\n",
    "        tf.add_to_collection('IG', activation)\n",
    "\n",
    "    cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=Y))\n",
    "\n",
    "    optimizer = tf.train.AdamOptimizer().minimize(cost, var_list=DNN.vars)\n",
    "\n",
    "    correct_prediction = tf.equal(tf.argmax(logits, 1), tf.argmax(Y, 1))\n",
    "    accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))\n",
    "\n",
    "cost_summary = tf.summary.scalar('Cost', cost)\n",
    "accuray_summary = tf.summary.scalar('Accuracy', accuracy)\n",
    "summary = tf.summary.merge_all()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. Training Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0001 cost = 0.163972841 accuracy = 0.948563643\n",
      "Epoch: 0002 cost = 0.042885760 accuracy = 0.986509101\n",
      "Epoch: 0003 cost = 0.028830733 accuracy = 0.991109099\n",
      "Epoch: 0004 cost = 0.023671029 accuracy = 0.992236371\n",
      "Epoch: 0005 cost = 0.017320657 accuracy = 0.994254551\n",
      "Epoch: 0006 cost = 0.013550124 accuracy = 0.995672731\n",
      "Epoch: 0007 cost = 0.013328977 accuracy = 0.995200005\n",
      "Epoch: 0008 cost = 0.008910857 accuracy = 0.997163639\n",
      "Epoch: 0009 cost = 0.010496597 accuracy = 0.996472731\n",
      "Epoch: 0010 cost = 0.007867829 accuracy = 0.997454548\n",
      "Epoch: 0011 cost = 0.007225713 accuracy = 0.997636366\n",
      "Epoch: 0012 cost = 0.008216114 accuracy = 0.997109094\n",
      "Epoch: 0013 cost = 0.006957658 accuracy = 0.997600002\n",
      "Epoch: 0014 cost = 0.004809637 accuracy = 0.997981820\n",
      "Epoch: 0015 cost = 0.005449639 accuracy = 0.998109093\n",
      "Accuracy: 0.9801\n"
     ]
    }
   ],
   "source": [
    "sess = tf.InteractiveSession()\n",
    "sess.run(tf.global_variables_initializer())\n",
    "\n",
    "saver = tf.train.Saver()\n",
    "file_writer = tf.summary.FileWriter(logdir, tf.get_default_graph())\n",
    "\n",
    "# Hyper parameters\n",
    "training_epochs = 15\n",
    "batch_size = 100\n",
    "\n",
    "for epoch in range(training_epochs):\n",
    "    total_batch = int(mnist.train.num_examples / batch_size)\n",
    "    avg_cost = 0\n",
    "    avg_acc = 0\n",
    "    \n",
    "    for i in range(total_batch):\n",
    "        batch_xs, batch_ys = mnist.train.next_batch(batch_size)\n",
    "        \n",
    "        # zero-center the images\n",
    "        mean = np.mean(batch_xs, axis=0, keepdims=True)\n",
    "        batch_xs = batch_xs - mean\n",
    "        \n",
    "        _, c, a, summary_str = sess.run([optimizer, cost, accuracy, summary], feed_dict={X: batch_xs, Y: batch_ys})\n",
    "        avg_cost += c / total_batch\n",
    "        avg_acc += a / total_batch\n",
    "        \n",
    "        file_writer.add_summary(summary_str, epoch * total_batch + i)\n",
    "\n",
    "    print('Epoch:', '%04d' % (epoch + 1), 'cost =', '{:.9f}'.format(avg_cost), 'accuracy =', '{:.9f}'.format(avg_acc))\n",
    "    \n",
    "    saver.save(sess, ckptdir)\n",
    "\n",
    "print('Accuracy:', sess.run(accuracy, feed_dict={X: mnist.test.images, Y: mnist.test.labels}))\n",
    "\n",
    "sess.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4. Restoring Graph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./tf_logs/3_4_IG/model\n"
     ]
    }
   ],
   "source": [
    "tf.reset_default_graph()\n",
    "\n",
    "sess = tf.InteractiveSession()\n",
    "\n",
    "new_saver = tf.train.import_meta_graph(ckptdir + '.meta')\n",
    "new_saver.restore(sess, tf.train.latest_checkpoint(logdir))\n",
    "\n",
    "activations = tf.get_collection('IG')\n",
    "X = activations[0]\n",
    "output = activations[-1]\n",
    "\n",
    "batch, _ = mnist.train.next_batch(100)\n",
    "mean = np.mean(batch, axis=0, keepdims=True)\n",
    "sample_imgs = [images[np.argmax(labels, axis=1) == i][1] - mean for i in range(10)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5. Displaying Images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAt0AAAQwCAYAAAA5PeKUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1e4328ab208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "steps = 100\n",
    "results = []\n",
    "\n",
    "for j in range(10):\n",
    "    gradient = tf.gradients(output[:,j], X)[0]\n",
    "    counterfs = [sample_imgs[j][0] * i / steps for i in range(steps + 1)]\n",
    "    grads = sess.run(gradient, feed_dict={X: counterfs})\n",
    "    integrad = np.zeros_like(sample_imgs[j])\n",
    "    \n",
    "    # This is same as counterfs[-1] * np.average(grads[1:], axis=0)\n",
    "    # I just used the for loop to make the Riemman Sum calculation step more explicit\n",
    "    for i in range(steps):\n",
    "        integrad += grads[i + 1] * (counterfs[i + 1] - counterfs[i])\n",
    "    \n",
    "    results.append(integrad)\n",
    "\n",
    "plt.figure(figsize=(15,15))\n",
    "for i in range(5):\n",
    "    plt.subplot(5, 2, 2 * i + 1)\n",
    "    plt.imshow(np.reshape(results[2 * i], [28, 28]), cmap='gray', interpolation='none')\n",
    "    plt.title('Digit: {}'.format(2 * i))\n",
    "    plt.xticks([])\n",
    "    plt.yticks([])\n",
    "    plt.colorbar()\n",
    "    \n",
    "    plt.subplot(5, 2, 2 * i + 2)\n",
    "    plt.imshow(np.reshape(results[2 * i + 1], [28, 28]), cmap='gray', interpolation='none')\n",
    "    plt.title('Digit: {}'.format(2 * i + 1))\n",
    "    plt.xticks([])\n",
    "    plt.yticks([])\n",
    "    plt.colorbar()\n",
    "\n",
    "plt.tight_layout()"
   ]
  },
  {
   "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"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autoclose": false,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}