{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2.1 Sensitivity Analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Introduction\n",
    "\n",
    "**This section corresponds to Section 4.1 of the original paper.**\n",
    "\n",
    "The task of explaining a DNN's decision can be rephrased into the following question: given a DNN $f$, input image $x$ and the output of DNN $f(x)$, what is it of $x$ that made the DNN reach the particular decision $f(x)$?\n",
    "\n",
    "To tackle this problem, we can define a relevance function $R$ that takes in an ith pixel $x_i$ of image $x$ as an argument and returns a scalar value $R(x_i)$ which indicates how much positive or negative contribution $x_i$ has had on the final decision $f(x)$. If the range of $R$ is limited to nonnegative real numbers, 0 indicates no contribution and any other positive scalar indicates positive contribution. Intuitively, the extent of positive/negative contribution must be proportional to the absolute value of $R(x_i)$.\n",
    "\n",
    "In this section, we ask for a given an image x, which pixel(s) or group(s) of pixels makes it representative of a certain concept encoded in the output neuron of the deep neural network (DNN).\n",
    "\n",
    "A first approach to identify the most important input features is sensitivity analysis. A common formulation of sensitivity analysis defines relevance scores as\n",
    "\n",
    "\\begin{equation}\n",
    "R(x_i) = \\left( \\frac{\\partial f}{\\partial x_i} \\right)^2\n",
    "\\end{equation}\n",
    "\n",
    "where the gradient is evaluated at the data point $x_i$. Put simply, the relevance score indicates how influential a particular data point was on the DNN's decision. Therefore, the most relevant input features are those to which the output is most sensitive i.e., has high relevance scores. The relevance score can be calculated simply using backpropagation.\n",
    "\n",
    "However, it is important to note that sensitivity analysis does not provide an explicit explanation of the function value $f(x)$, but rather suggestions. For instance, when applying sensitivity analysis to a DNN detecting cars, we answer the question *\"what makes this image more/less a car\"* rather than the more basic question *\"what makes this image a car?\"*. This may not make much sense to you. Let me offer an explanation with a simplified version of sensitivity analysis.\n",
    "\n",
    "Suppose we have a binary classification function that separates 2-dimensional points into two classes:\n",
    "\n",
    "\\begin{equation}\n",
    "f(x,y) = \n",
    "\\begin{cases}\n",
    "x^2 + y^2 - 9 < 0 & \\text{for class N(egative)} \\\\\n",
    "x^2 + y^2 - 9 \\geq 0 &  \\text{for class P(ositive)}\n",
    "\\end{cases}\n",
    "\\end{equation}\n",
    "\n",
    "For instance, $f(-1,1) = -7 < 0$ and therefore the point $(-1,1)$ is in class $N$. On the other hand, $f(-4,-4) = 7 > 0$ and therefore the point $(-4,-4)$ is in class $P$.\n",
    "\n",
    "Let us now turn our attention to the gradient of $f$ at points $(-1,1)$ and $(-4,-4)$. The gradient of $f$ is:\n",
    "\n",
    "\\begin{equation}\n",
    "\\nabla f(x,y) = \n",
    "\\begin{bmatrix}\n",
    "\\partial f / \\partial x & \\partial f / \\partial y\n",
    "\\end{bmatrix} = \n",
    "\\begin{bmatrix}\n",
    "2x & 2y\n",
    "\\end{bmatrix}\n",
    "\\end{equation}\n",
    "\n",
    "The gradient at $(-1,1)$ is given by $\\begin{bmatrix}-2 & 2\\end{bmatrix}$ and the sensitivity is $\\begin{bmatrix}4 & 4\\end{bmatrix}$. Note that the sensitivity does not tell *why* $(-1,1)$ is in class $N$. It merely suggests that increasing or decreasing $x$ and $y$ may produce a point that belongs to class $N$ more strongly.\n",
    "\n",
    "Let's take a look at another example: $(-4,-4)$. The gradient at the point is given by $\\begin{bmatrix}-8 & -8\\end{bmatrix}$ and therefore the sensitivity is $\\begin{bmatrix}64 & 64\\end{bmatrix}$. Also in this case, the sensitivity only suggests that increasing or decreasing $x$ and $y$ may produce a point that DNN will classify into class $P$ with higher confidence (see the figure below for a visualization of the points with their normalized gradients indicated by arrows).\n",
    "\n",
    "Likewise, the sensitivity scores for images do not tell us *why* the DNN made such a decision. It only tells us what changes in the image (e.g. emphasis or de-emphasis of a particular feature) may cause the DNN to classify the image with more certainty into the given class. Looking back, note how apt the name *\"sensitivity analysis\"* is. Sensitivity analysis tells us how *sensitive* the function value $f(x)$ is to changes in pixel values. The magnitude of sensitivity is quantified by the relevance score $R(x_i)$ given by the square of gradient of $f(x)$ at the pixel $x_i$.\n",
    "\n",
    "![title](./assets/2_1_SA/contour.png)"
   ]
  },
  {
   "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",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "\n",
    "from models.models_2_1 import MNIST_DNN, MNIST_CNN\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "mnist = input_data.read_data_sets(\"MNIST_data/\", one_hot=True)\n",
    "\n",
    "logdir = './tf_logs/2_1_SA/'\n",
    "ckptdir = logdir + 'model'\n",
    "\n",
    "if not os.path.exists(logdir):\n",
    "    os.mkdir(logdir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. Building Graph\n",
    "\n",
    "In this step, a CNN classifier is initialized and necessary nodes for model training is added onto the graph. For sensitivity analysis, we only need three placeholders. `training` to indicate that we do not use dropout when during inference, `X` in order to pass test images into the computation graph and `logits` which is necessary to compute the gradient (we need to compute the gradient of `X` with respect to `logits`). Instead of using a fully connected neural network, I used a convolutional neural network because it provided better sensitivity analysis visualization."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "with tf.name_scope('Classifier'):\n",
    "\n",
    "    # Initialize neural network\n",
    "    CNN = MNIST_CNN('CNN')\n",
    "\n",
    "    # Setup training process\n",
    "    training = tf.placeholder_with_default(True, shape=[], name='training')\n",
    "    X = tf.placeholder(tf.float32, [None, 784], name='X')\n",
    "    Y = tf.placeholder(tf.float32, [None, 10], name='Y')\n",
    "\n",
    "    logits = CNN(X, training)\n",
    "    \n",
    "    tf.add_to_collection('sensitivity_analysis', training)\n",
    "    tf.add_to_collection('sensitivity_analysis', X)\n",
    "    tf.add_to_collection('sensitivity_analysis', logits)\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=CNN.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\n",
    "\n",
    "This is the step where the CNN is trained to classify the 10 digits of the MNIST images. Summaries are written into the logdir and you can visualize the statistics using tensorboard by typing this command: `tensorboard --lodir=./tf_logs`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0001 cost = 0.755203515 accuracy = 0.744127272\n",
      "Epoch: 0002 cost = 0.287379274 accuracy = 0.909418183\n",
      "Epoch: 0003 cost = 0.230599659 accuracy = 0.927818183\n",
      "Epoch: 0004 cost = 0.198606952 accuracy = 0.937472728\n",
      "Epoch: 0005 cost = 0.183995720 accuracy = 0.942200003\n",
      "Epoch: 0006 cost = 0.170081426 accuracy = 0.946763638\n",
      "Epoch: 0007 cost = 0.165785470 accuracy = 0.948218184\n",
      "Epoch: 0008 cost = 0.155800372 accuracy = 0.951472731\n",
      "Epoch: 0009 cost = 0.151487274 accuracy = 0.952563640\n",
      "Epoch: 0010 cost = 0.145716070 accuracy = 0.954636367\n",
      "Epoch: 0011 cost = 0.142153567 accuracy = 0.955272730\n",
      "Epoch: 0012 cost = 0.138724105 accuracy = 0.957436367\n",
      "Epoch: 0013 cost = 0.140947498 accuracy = 0.957090912\n",
      "Epoch: 0014 cost = 0.137629085 accuracy = 0.957836368\n",
      "Epoch: 0015 cost = 0.136273555 accuracy = 0.958454548\n",
      "Accuracy: 0.9892\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",
    "        _, 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, training: False}))\n",
    "\n",
    "sess.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4. Restoring Subgraph\n",
    "\n",
    "Here we first rebuild the CNN graph from metagraph, restore CNN parameters from the checkpoint and then gather the necessary nodes for Sensitivity Analysis using the `tf.get_collection()` function. We can calculate the relevance scores using the `tf.square()` and the `tf.gradients()` functions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./tf_logs/2_1_SA/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",
    "SA = tf.get_collection('sensitivity_analysis')\n",
    "training = SA[0]\n",
    "X = SA[1]\n",
    "logits = SA[2]\n",
    "\n",
    "SA_scores = [tf.square(tf.gradients(logits[:,i], X)) for i in range(10)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5. Calculating Relevance Scores $R_i(x)$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "images = mnist.train.images\n",
    "labels = mnist.train.labels\n",
    "\n",
    "sample_imgs = []\n",
    "for i in range(10):\n",
    "    sample_imgs.append(images[np.argmax(labels, axis=1) == i][3])\n",
    "\n",
    "hmaps = np.reshape([sess.run(SA_scores[i], feed_dict={X: sample_imgs[i][None,:], training:False}) for i in range(10)], [10, 784])\n",
    "\n",
    "sess.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6. Displaying Images\n",
    "\n",
    "The relevance scores are visualized as heat maps over the original digit images. You can see which features/data points influenced the CNN most its decision making."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtEAAAQwCAYAAAAjAQIaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a6ce0ee278>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15,15))\n",
    "for i in range(5):\n",
    "    plt.subplot(5, 2, 2 * i + 1)\n",
    "    plt.imshow(np.reshape(sample_imgs[2 * i], [28, 28]), cmap='gray')\n",
    "    plt.imshow(np.reshape(hmaps[2 * i], [28, 28]), cmap='hot', alpha=0.5)\n",
    "    plt.title('Digit: {}'.format(2 * i))\n",
    "    plt.colorbar()\n",
    "    \n",
    "    plt.subplot(5, 2, 2 * i + 2)\n",
    "    plt.imshow(np.reshape(sample_imgs[2 * i + 1], [28, 28]), cmap='gray')\n",
    "    plt.imshow(np.reshape(hmaps[2 * i + 1], [28, 28]), cmap='hot', alpha=0.5)\n",
    "    plt.title('Digit: {}'.format(2 * i + 1))\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
}