part8-9-ImageRecognition.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    }
   ],
   "source": [
    "%pylab inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# http://www.bogotobogo.com/python/scikit-learn/Artificial-Neural-Network-ANN-8-Deep-Learning-1-Image-Recognition-Image-Uploading.php"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import struct\n",
    "import numpy as np\n",
    "\n",
    "def load_mnist(path, kind='train'):\n",
    "    \"\"\"Load MNIST data from `path`\"\"\"\n",
    "    labels_path = os.path.join(path, '%s-labels-idx1-ubyte' % kind)\n",
    "    images_path = os.path.join(path, '%s-images-idx3-ubyte' % kind)\n",
    "        \n",
    "    with open(labels_path, 'rb') as lbpath:\n",
    "        magic, n = struct.unpack('>II', lbpath.read(8))\n",
    "        labels = np.fromfile(lbpath, dtype=np.uint8)\n",
    "\n",
    "    with open(images_path, 'rb') as imgpath:\n",
    "        magic, num, rows, cols = struct.unpack(\">IIII\", imgpath.read(16))\n",
    "        images = np.fromfile(imgpath, dtype=np.uint8).reshape(len(labels), 784)\n",
    " \n",
    "    return images, labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 784)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train, y_train = load_mnist('mnist/', kind='train')\n",
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10000, 784)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test, y_test = load_mnist('mnist', kind='t10k')\n",
    "X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAk4AAAGGCAYAAACNCg6xAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f487ada9dd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(nrows=2, ncols=5, sharex=True, sharey=True)\n",
    "ax = ax.flatten()\n",
    "for i in range(10):\n",
    "    img = X_train[y_train == i][0].reshape(28, 28)\n",
    "    ax[i].imshow(img, cmap='Greys', interpolation='nearest')\n",
    "ax[0].set_xticks([])\n",
    "ax[0].set_yticks([])\n",
    "tight_layout()\n",
    "show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAk4AAAGGCAYAAACNCg6xAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4879729550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(nrows=5, ncols=5, sharex=True, sharey=True,)\n",
    "ax = ax.flatten()\n",
    "for i in range(25):\n",
    "    img = X_train[y_train == 4][i].reshape(28, 28)\n",
    "    ax[i].imshow(img, cmap='Greys', interpolation='nearest')\n",
    "ax[0].set_xticks([])\n",
    "ax[0].set_yticks([])\n",
    "tight_layout()\n",
    "show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAk4AAAGGCAYAAACNCg6xAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4878edc3d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(nrows=5, ncols=5, sharex=True, sharey=True,)\n",
    "ax = ax.flatten()\n",
    "for i in range(25):\n",
    "    img = X_train[y_train == 7][i].reshape(28, 28)\n",
    "    ax[i].imshow(img, cmap='Greys', interpolation='nearest')\n",
    "ax[0].set_xticks([])\n",
    "ax[0].set_yticks([])\n",
    "tight_layout()\n",
    "show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAk4AAAGGCAYAAACNCg6xAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4879ae0590>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(nrows=5, ncols=5, sharex=True, sharey=True,)\n",
    "ax = ax.flatten()\n",
    "for i in range(25):\n",
    "    img = X_train[y_train == 9][i].reshape(28, 28)\n",
    "    ax[i].imshow(img, cmap='Greys', interpolation='nearest')\n",
    "ax[0].set_xticks([])\n",
    "ax[0].set_yticks([])\n",
    "tight_layout()\n",
    "show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# http://www.bogotobogo.com/python/scikit-learn/Neural-Network-ANN-9-Deep-Learning-2-Image-Recognition-Image-Classification.php"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from scipy.special import expit\n",
    "import sys\n",
    "\n",
    "\n",
    "class NeuralNetMLP(object):\n",
    "    \"\"\" Feedforward neural network / Multi-layer perceptron classifier.\n",
    "\n",
    "    Parameters\n",
    "    ------------\n",
    "    n_output : int\n",
    "        Number of output units, should be equal to the\n",
    "        number of unique class labels.\n",
    "    n_features : int\n",
    "        Number of features (dimensions) in the target dataset.\n",
    "        Should be equal to the number of columns in the X array.\n",
    "    n_hidden : int (default: 30)\n",
    "        Number of hidden units.\n",
    "    l1 : float (default: 0.0)\n",
    "        Lambda value for L1-regularization.\n",
    "        No regularization if l1=0.0 (default)\n",
    "    l2 : float (default: 0.0)\n",
    "        Lambda value for L2-regularization.\n",
    "        No regularization if l2=0.0 (default)\n",
    "    epochs : int (default: 500)\n",
    "        Number of passes over the training set.\n",
    "    eta : float (default: 0.001)\n",
    "        Learning rate.\n",
    "    alpha : float (default: 0.0)\n",
    "        Momentum constant. Factor multiplied with the\n",
    "        gradient of the previous epoch t-1 to improve\n",
    "        learning speed\n",
    "        w(t) := w(t) - (grad(t) + alpha*grad(t-1))\n",
    "    decrease_const : float (default: 0.0)\n",
    "        Decrease constant. Shrinks the learning rate\n",
    "        after each epoch via eta / (1 + epoch*decrease_const)\n",
    "    shuffle : bool (default: True)\n",
    "        Shuffles training data every epoch if True to prevent circles.\n",
    "    minibatches : int (default: 1)\n",
    "        Divides training data into k minibatches for efficiency.\n",
    "        Normal gradient descent learning if k=1 (default).\n",
    "    random_state : int (default: None)\n",
    "        Set random state for shuffling and initializing the weights.\n",
    "\n",
    "    Attributes\n",
    "    -----------\n",
    "    cost_ : list\n",
    "      Sum of squared errors after each epoch.\n",
    "\n",
    "    \"\"\"\n",
    "    def __init__(self, n_output, n_features, n_hidden=30,\n",
    "                 l1=0.0, l2=0.0, epochs=500, eta=0.001,\n",
    "                 alpha=0.0, decrease_const=0.0, shuffle=True,\n",
    "                 minibatches=1, random_state=None):\n",
    "\n",
    "        np.random.seed(random_state)\n",
    "        self.n_output = n_output\n",
    "        self.n_features = n_features\n",
    "        self.n_hidden = n_hidden\n",
    "        self.w1, self.w2 = self._initialize_weights()\n",
    "        self.l1 = l1\n",
    "        self.l2 = l2\n",
    "        self.epochs = epochs\n",
    "        self.eta = eta\n",
    "        self.alpha = alpha\n",
    "        self.decrease_const = decrease_const\n",
    "        self.shuffle = shuffle\n",
    "        self.minibatches = minibatches\n",
    "\n",
    "    def _encode_labels(self, y, k):\n",
    "        \"\"\"Encode labels into one-hot representation\n",
    "\n",
    "        Parameters\n",
    "        ------------\n",
    "        y : array, shape = [n_samples]\n",
    "            Target values.\n",
    "\n",
    "        Returns\n",
    "        -----------\n",
    "        onehot : array, shape = (n_labels, n_samples)\n",
    "\n",
    "        \"\"\"\n",
    "        onehot = np.zeros((k, y.shape[0]))\n",
    "        for idx, val in enumerate(y):\n",
    "            onehot[val, idx] = 1.0\n",
    "        return onehot\n",
    "\n",
    "    def _initialize_weights(self):\n",
    "        \"\"\"Initialize weights with small random numbers.\"\"\"\n",
    "        w1 = np.random.uniform(-1.0, 1.0,\n",
    "                               size=self.n_hidden*(self.n_features + 1))\n",
    "        w1 = w1.reshape(self.n_hidden, self.n_features + 1)\n",
    "        w2 = np.random.uniform(-1.0, 1.0,\n",
    "                               size=self.n_output*(self.n_hidden + 1))\n",
    "        w2 = w2.reshape(self.n_output, self.n_hidden + 1)\n",
    "        return w1, w2\n",
    "\n",
    "    def _sigmoid(self, z):\n",
    "        \"\"\"Compute logistic function (sigmoid)\n",
    "\n",
    "        Uses scipy.special.expit to avoid overflow\n",
    "        error for very small input values z.\n",
    "\n",
    "        \"\"\"\n",
    "        # return 1.0 / (1.0 + np.exp(-z))\n",
    "        return expit(z)\n",
    "\n",
    "    def _sigmoid_gradient(self, z):\n",
    "        \"\"\"Compute gradient of the logistic function\"\"\"\n",
    "        sg = self._sigmoid(z)\n",
    "        return sg * (1.0 - sg)\n",
    "\n",
    "    def _add_bias_unit(self, X, how='column'):\n",
    "        \"\"\"Add bias unit (column or row of 1s) to array at index 0\"\"\"\n",
    "        if how == 'column':\n",
    "            X_new = np.ones((X.shape[0], X.shape[1] + 1))\n",
    "            X_new[:, 1:] = X\n",
    "        elif how == 'row':\n",
    "            X_new = np.ones((X.shape[0] + 1, X.shape[1]))\n",
    "            X_new[1:, :] = X\n",
    "        else:\n",
    "            raise AttributeError('`how` must be `column` or `row`')\n",
    "        return X_new\n",
    "\n",
    "    def _feedforward(self, X, w1, w2):\n",
    "        \"\"\"Compute feedforward step\n",
    "\n",
    "        Parameters\n",
    "        -----------\n",
    "        X : array, shape = [n_samples, n_features]\n",
    "            Input layer with original features.\n",
    "        w1 : array, shape = [n_hidden_units, n_features]\n",
    "            Weight matrix for input layer -> hidden layer.\n",
    "        w2 : array, shape = [n_output_units, n_hidden_units]\n",
    "            Weight matrix for hidden layer -> output layer.\n",
    "\n",
    "        Returns\n",
    "        ----------\n",
    "        a1 : array, shape = [n_samples, n_features+1]\n",
    "            Input values with bias unit.\n",
    "        z2 : array, shape = [n_hidden, n_samples]\n",
    "            Net input of hidden layer.\n",
    "        a2 : array, shape = [n_hidden+1, n_samples]\n",
    "            Activation of hidden layer.\n",
    "        z3 : array, shape = [n_output_units, n_samples]\n",
    "            Net input of output layer.\n",
    "        a3 : array, shape = [n_output_units, n_samples]\n",
    "            Activation of output layer.\n",
    "\n",
    "        \"\"\"\n",
    "        a1 = self._add_bias_unit(X, how='column')\n",
    "        z2 = w1.dot(a1.T)\n",
    "        a2 = self._sigmoid(z2)\n",
    "        a2 = self._add_bias_unit(a2, how='row')\n",
    "        z3 = w2.dot(a2)\n",
    "        a3 = self._sigmoid(z3)\n",
    "        return a1, z2, a2, z3, a3\n",
    "\n",
    "    def _L2_reg(self, lambda_, w1, w2):\n",
    "        \"\"\"Compute L2-regularization cost\"\"\"\n",
    "        return (lambda_/2.0) * (np.sum(w1[:, 1:] ** 2) +\n",
    "                                np.sum(w2[:, 1:] ** 2))\n",
    "\n",
    "    def _L1_reg(self, lambda_, w1, w2):\n",
    "        \"\"\"Compute L1-regularization cost\"\"\"\n",
    "        return (lambda_/2.0) * (np.abs(w1[:, 1:]).sum() +\n",
    "                                np.abs(w2[:, 1:]).sum())\n",
    "\n",
    "    def _get_cost(self, y_enc, output, w1, w2):\n",
    "        \"\"\"Compute cost function.\n",
    "\n",
    "        Parameters\n",
    "        ----------\n",
    "        y_enc : array, shape = (n_labels, n_samples)\n",
    "            one-hot encoded class labels.\n",
    "        output : array, shape = [n_output_units, n_samples]\n",
    "            Activation of the output layer (feedforward)\n",
    "        w1 : array, shape = [n_hidden_units, n_features]\n",
    "            Weight matrix for input layer -> hidden layer.\n",
    "        w2 : array, shape = [n_output_units, n_hidden_units]\n",
    "            Weight matrix for hidden layer -> output layer.\n",
    "\n",
    "        Returns\n",
    "        ---------\n",
    "        cost : float\n",
    "            Regularized cost.\n",
    "\n",
    "        \"\"\"\n",
    "        term1 = -y_enc * (np.log(output))\n",
    "        term2 = (1.0 - y_enc) * np.log(1.0 - output)\n",
    "        cost = np.sum(term1 - term2)\n",
    "        L1_term = self._L1_reg(self.l1, w1, w2)\n",
    "        L2_term = self._L2_reg(self.l2, w1, w2)\n",
    "        cost = cost + L1_term + L2_term\n",
    "        return cost\n",
    "\n",
    "    def _get_gradient(self, a1, a2, a3, z2, y_enc, w1, w2):\n",
    "        \"\"\" Compute gradient step using backpropagation.\n",
    "\n",
    "        Parameters\n",
    "        ------------\n",
    "        a1 : array, shape = [n_samples, n_features+1]\n",
    "            Input values with bias unit.\n",
    "        a2 : array, shape = [n_hidden+1, n_samples]\n",
    "            Activation of hidden layer.\n",
    "        a3 : array, shape = [n_output_units, n_samples]\n",
    "            Activation of output layer.\n",
    "        z2 : array, shape = [n_hidden, n_samples]\n",
    "            Net input of hidden layer.\n",
    "        y_enc : array, shape = (n_labels, n_samples)\n",
    "            one-hot encoded class labels.\n",
    "        w1 : array, shape = [n_hidden_units, n_features]\n",
    "            Weight matrix for input layer -> hidden layer.\n",
    "        w2 : array, shape = [n_output_units, n_hidden_units]\n",
    "            Weight matrix for hidden layer -> output layer.\n",
    "\n",
    "        Returns\n",
    "        ---------\n",
    "        grad1 : array, shape = [n_hidden_units, n_features]\n",
    "            Gradient of the weight matrix w1.\n",
    "        grad2 : array, shape = [n_output_units, n_hidden_units]\n",
    "            Gradient of the weight matrix w2.\n",
    "\n",
    "        \"\"\"\n",
    "        # backpropagation\n",
    "        sigma3 = a3 - y_enc\n",
    "        z2 = self._add_bias_unit(z2, how='row')\n",
    "        sigma2 = w2.T.dot(sigma3) * self._sigmoid_gradient(z2)\n",
    "        sigma2 = sigma2[1:, :]\n",
    "        grad1 = sigma2.dot(a1)\n",
    "        grad2 = sigma3.dot(a2.T)\n",
    "\n",
    "        # regularize\n",
    "        grad1[:, 1:] += self.l2 * w1[:, 1:]\n",
    "        grad1[:, 1:] += self.l1 * np.sign(w1[:, 1:])\n",
    "        grad2[:, 1:] += self.l2 * w2[:, 1:]\n",
    "        grad2[:, 1:] += self.l1 * np.sign(w2[:, 1:])\n",
    "\n",
    "        return grad1, grad2\n",
    "\n",
    "    def predict(self, X):\n",
    "        \"\"\"Predict class labels\n",
    "\n",
    "        Parameters\n",
    "        -----------\n",
    "        X : array, shape = [n_samples, n_features]\n",
    "            Input layer with original features.\n",
    "\n",
    "        Returns:\n",
    "        ----------\n",
    "        y_pred : array, shape = [n_samples]\n",
    "            Predicted class labels.\n",
    "\n",
    "        \"\"\"\n",
    "        if len(X.shape) != 2:\n",
    "            raise AttributeError('X must be a [n_samples, n_features] array.\\n'\n",
    "                                 'Use X[:,None] for 1-feature classification,'\n",
    "                                 '\\nor X[[i]] for 1-sample classification')\n",
    "\n",
    "        a1, z2, a2, z3, a3 = self._feedforward(X, self.w1, self.w2)\n",
    "        y_pred = np.argmax(z3, axis=0)\n",
    "        return y_pred\n",
    "\n",
    "    def fit(self, X, y, print_progress=False):\n",
    "        \"\"\" Learn weights from training data.\n",
    "\n",
    "        Parameters\n",
    "        -----------\n",
    "        X : array, shape = [n_samples, n_features]\n",
    "            Input layer with original features.\n",
    "        y : array, shape = [n_samples]\n",
    "            Target class labels.\n",
    "        print_progress : bool (default: False)\n",
    "            Prints progress as the number of epochs\n",
    "            to stderr.\n",
    "\n",
    "        Returns:\n",
    "        ----------\n",
    "        self\n",
    "\n",
    "        \"\"\"\n",
    "        self.cost_ = []\n",
    "        X_data, y_data = X.copy(), y.copy()\n",
    "        y_enc = self._encode_labels(y, self.n_output)\n",
    "\n",
    "        delta_w1_prev = np.zeros(self.w1.shape)\n",
    "        delta_w2_prev = np.zeros(self.w2.shape)\n",
    "\n",
    "        for i in range(self.epochs):\n",
    "\n",
    "            # adaptive learning rate\n",
    "            self.eta /= (1 + self.decrease_const*i)\n",
    "\n",
    "            if print_progress:\n",
    "                sys.stderr.write('\\rEpoch: %d/%d' % (i+1, self.epochs))\n",
    "                sys.stderr.flush()\n",
    "\n",
    "            if self.shuffle:\n",
    "                idx = np.random.permutation(y_data.shape[0])\n",
    "                X_data, y_enc = X_data[idx], y_enc[:, idx]\n",
    "\n",
    "            mini = np.array_split(range(y_data.shape[0]), self.minibatches)\n",
    "            for idx in mini:\n",
    "\n",
    "                # feedforward\n",
    "                a1, z2, a2, z3, a3 = self._feedforward(X_data[idx],\n",
    "                                                       self.w1,\n",
    "                                                       self.w2)\n",
    "                cost = self._get_cost(y_enc=y_enc[:, idx],\n",
    "                                      output=a3,\n",
    "                                      w1=self.w1,\n",
    "                                      w2=self.w2)\n",
    "                self.cost_.append(cost)\n",
    "\n",
    "                # compute gradient via backpropagation\n",
    "                grad1, grad2 = self._get_gradient(a1=a1, a2=a2,\n",
    "                                                  a3=a3, z2=z2,\n",
    "                                                  y_enc=y_enc[:, idx],\n",
    "                                                  w1=self.w1,\n",
    "                                                  w2=self.w2)\n",
    "\n",
    "                delta_w1, delta_w2 = self.eta * grad1, self.eta * grad2\n",
    "                self.w1 -= (delta_w1 + (self.alpha * delta_w1_prev))\n",
    "                self.w2 -= (delta_w2 + (self.alpha * delta_w2_prev))\n",
    "                delta_w1_prev, delta_w2_prev = delta_w1, delta_w2\n",
    "\n",
    "        return self"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<__main__.NeuralNetMLP at 0x7f4879ff8910>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# initialization\n",
    "nn = NeuralNetMLP(n_output=10, \n",
    "                  n_features=X_train.shape[1], \n",
    "                  n_hidden=50, \n",
    "                  l2=0.1, \n",
    "                  l1=0.0, \n",
    "                  epochs=1000, \n",
    "                  eta=0.001,\n",
    "                  alpha=0.001,\n",
    "                  decrease_const=0.00001,\n",
    "                  minibatches=50, \n",
    "                  shuffle=True,\n",
    "                  random_state=1)\n",
    "nn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch: 1000/1000"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<__main__.NeuralNetMLP at 0x7f4879ff8910>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# training\n",
    "nn.fit(X_train, y_train, print_progress=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAk4AAAGGCAYAAACNCg6xAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f486e5a2710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot(range(len(nn.cost_)), nn.cost_, color='k')\n",
    "ylim([0, 2000])\n",
    "ylabel('Cost')\n",
    "xlabel('Epochs * 50')\n",
    "tight_layout()\n",
    "show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "batches = np.array_split(range(len(nn.cost_)), 1000)\n",
    "cost_array = np.array(nn.cost_)\n",
    "cost_averages = [np.mean(cost_array[i]) for i in batches]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAk4AAAGGCAYAAACNCg6xAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJEmqkMFJkiSpQv8f1dXUUytRnpgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f486e54fa90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot(range(len(cost_averages)), cost_averages, color='b')\n",
    "ylim([0, 2000])\n",
    "ylabel('Cost')\n",
    "xlabel('Epochs')\n",
    "tight_layout()\n",
    "show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.97571666666666668"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train_pred = nn.predict(X_train)\n",
    "accuracy = \\\n",
    "  ((np.sum(y_train == y_train_pred, axis=0)).astype('float') / X_train.shape[0])\n",
    "accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.95679999999999998"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test_pred = nn.predict(X_test)\n",
    "accuracy = \\\n",
    "  ((np.sum(y_test == y_test_pred, axis=0)).astype('float') / X_test.shape[0])\n",
    "accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkoAAAGGCAYAAACE4a7LAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\noiQIgiAIghCA/wfR0OTLoL56igAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f487b105a50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "misclassified_img = X_test[y_test != y_test_pred][:30]\n",
    "correct_lab = y_test[y_test != y_test_pred][:30]\n",
    "misclassified_lab = y_test_pred[y_test != y_test_pred][:30]\n",
    "\n",
    "fig, ax = plt.subplots(nrows=6, ncols=5, sharex=True, sharey=True,)\n",
    "ax = ax.flatten()\n",
    "for i in range(30):\n",
    "    img = misclassified_img[i].reshape(28, 28)\n",
    "    ax[i].imshow(img, cmap='Greys', interpolation='nearest')\n",
    "    ax[i].set_title('%d) t: %d p: %d' % (i+1, correct_lab[i], misclassified_lab[i]))\n",
    "\n",
    "ax[0].set_xticks([])\n",
    "ax[0].set_yticks([])\n",
    "tight_layout()\n",
    "show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}