Convolutional Neural Network.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\h5py\\__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n",
      "Using Theano backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.datasets import fashion_mnist\n",
    "(train_X, train_Y), (test_X, test_Y) = fashion_mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training data shape: (60000, 28, 28) (60000,)\n",
      "Testing data shape: (10000, 28, 28) (10000,)\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "from keras.utils import to_categorical\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "print('Training data shape:', train_X.shape, train_Y.shape)\n",
    "print('Testing data shape:', test_X.shape, test_Y.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total no of outputs 10\n",
      "Output classes [0 1 2 3 4 5 6 7 8 9]\n"
     ]
    }
   ],
   "source": [
    "classes = np.unique(train_Y)\n",
    "nClasses = len(classes)\n",
    "print('Total no of outputs', nClasses)\n",
    "print('Output classes', classes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,1,'Ground Truth: 9')"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f350c392e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=[6,6])\n",
    "#dispalying the first train image\n",
    "plt.subplot(121)\n",
    "plt.imshow(train_X[0,:,:], cmap='gray')\n",
    "plt.title(\"Ground Truth: {}\".format(train_Y[0]))\n",
    "#displaying the first test image\n",
    "plt.subplot(122)\n",
    "plt.imshow(test_X[0,:,:], cmap='gray')\n",
    "plt.title(\"Ground Truth: {}\".format(test_Y[0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data Processing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((60000, 28, 28, 1), (10000, 28, 28, 1))"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_X =train_X.reshape(-1,28,28,1)\n",
    "test_X =test_X.reshape(-1,28,28,1)\n",
    "train_X.shape, test_X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_X=train_X.astype('float32')\n",
    "test_X=test_X.astype('float32')\n",
    "\n",
    "train_X = train_X / 255\n",
    "test_X= test_X / 255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original Label: 9\n",
      "After conversion to one-hot: [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n"
     ]
    }
   ],
   "source": [
    "train_Y_one_hot = to_categorical(train_Y)\n",
    "test_Y_one_hot = to_categorical(test_Y)\n",
    "\n",
    "print('Original Label:', train_Y[0])\n",
    "print('After conversion to one-hot:', train_Y_one_hot[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "train_X,valid_X,train_label,valid_label = train_test_split(train_X,train_Y_one_hot, test_size=0.2, random_state=13)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# model the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import keras\n",
    "from keras.models import Sequential, Input, Model\n",
    "from keras.layers import Dense, Dropout, Flatten\n",
    "from keras.layers import Conv2D, MaxPooling2D\n",
    "from keras.layers.normalization import BatchNormalization\n",
    "from keras.layers.advanced_activations import LeakyReLU\n",
    "\n",
    "batch_size =64\n",
    "epochs = 2\n",
    "num_classes =10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "fashion_model = Sequential()\n",
    "fashion_model.add(Conv2D(32, kernel_size=(3, 3), activation ='linear',data_format=\"channels_last\",input_shape=(28,28,1), padding='same'))\n",
    "fashion_model.add(LeakyReLU(alpha=0.1))\n",
    "fashion_model.add(MaxPooling2D((2,2), padding='same'))\n",
    "fashion_model.add(Conv2D(64, (3,3), activation ='linear', padding ='same'))\n",
    "fashion_model.add(LeakyReLU(alpha=0.1))\n",
    "fashion_model.add(MaxPooling2D(pool_size=(2,2), padding='same'))\n",
    "fashion_model.add(Conv2D(128, (3,3), activation='linear', padding='same'))\n",
    "fashion_model.add(LeakyReLU(alpha=0.1))\n",
    "fashion_model.add(MaxPooling2D(pool_size=(2,2),padding='same'))\n",
    "fashion_model.add(Flatten())\n",
    "fashion_model.add(Dense(128, activation='linear'))\n",
    "fashion_model.add(LeakyReLU(alpha=0.1))\n",
    "fashion_model.add(Dense(num_classes,activation='softmax'))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "fashion_model.compile(loss=keras.losses.categorical_crossentropy, optimizer=keras.optimizers.Adam(),metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([[[[-0.09945804,  0.0915065 ,  0.03485741,  0.08161819,\n",
       "            0.08119676,  0.07725154, -0.07746923, -0.0223191 ,\n",
       "           -0.05162062,  0.08703659,  0.04245423,  0.06205128,\n",
       "           -0.08285289, -0.09201675, -0.03784846,  0.13279815,\n",
       "            0.07207875,  0.04125308,  0.0090504 ,  0.0620718 ,\n",
       "           -0.02916222,  0.01322412, -0.00527263, -0.08186387,\n",
       "           -0.08391537, -0.10227016, -0.06112178, -0.08593874,\n",
       "            0.05368049,  0.05811486,  0.01558812,  0.01546992]],\n",
       " \n",
       "         [[-0.04137689,  0.0308729 , -0.08471477,  0.09032054,\n",
       "           -0.10671425, -0.12254865,  0.08711937,  0.12238993,\n",
       "            0.0037889 , -0.04131994,  0.04753834, -0.0656004 ,\n",
       "            0.08671738,  0.06353861,  0.09234776, -0.05339721,\n",
       "            0.00402781,  0.07460652, -0.06749061,  0.01039584,\n",
       "           -0.06269942, -0.05914698, -0.04248975,  0.01445159,\n",
       "           -0.0810686 , -0.02815664,  0.13802241, -0.13463986,\n",
       "           -0.05177849,  0.05462985, -0.07831627, -0.0373854 ]],\n",
       " \n",
       "         [[ 0.10120857, -0.05595601, -0.09047458, -0.12507756,\n",
       "           -0.10123153,  0.05401476,  0.08053511, -0.12788211,\n",
       "            0.0168844 , -0.07681486, -0.01785665,  0.04738487,\n",
       "           -0.05941412, -0.07878397, -0.11538964,  0.13698946,\n",
       "            0.11122657,  0.10447985, -0.07632389, -0.07568075,\n",
       "            0.13783465, -0.06521254, -0.07275087,  0.01431397,\n",
       "           -0.04614972,  0.01953591,  0.07980369,  0.06154512,\n",
       "            0.02423182, -0.04115222, -0.1169277 ,  0.08425424]]],\n",
       " \n",
       " \n",
       "        [[[-0.08584376, -0.00368258,  0.06804898, -0.04272546,\n",
       "           -0.1313429 , -0.03350981,  0.01296689,  0.02245896,\n",
       "           -0.11227748,  0.02677356,  0.0353359 ,  0.07053527,\n",
       "           -0.01405425,  0.00114265, -0.01670733, -0.06926099,\n",
       "           -0.13903858, -0.12788086,  0.13828842,  0.0927572 ,\n",
       "           -0.11113721,  0.10790204,  0.05482642, -0.00261877,\n",
       "           -0.04416109, -0.12705427, -0.02052067,  0.1249605 ,\n",
       "            0.00038075, -0.12081591, -0.06934853, -0.0222474 ]],\n",
       " \n",
       "         [[-0.07217614,  0.12194152,  0.1214648 ,  0.04753682,\n",
       "           -0.02703459,  0.13287638, -0.12359631, -0.07794374,\n",
       "           -0.10290596, -0.11753848, -0.02168737, -0.06047232,\n",
       "           -0.14178467,  0.11374386,  0.04309477,  0.13823415,\n",
       "           -0.02037293, -0.09032327, -0.03528399, -0.10412034,\n",
       "           -0.10006835,  0.06402774,  0.07136032,  0.01609094,\n",
       "            0.04942331,  0.04800197, -0.09057578, -0.12717612,\n",
       "           -0.12418555,  0.10233535,  0.06279077,  0.14168416]],\n",
       " \n",
       "         [[ 0.11133052,  0.10320017,  0.10073055,  0.00388789,\n",
       "            0.11269052, -0.11981533,  0.04668526, -0.04328858,\n",
       "           -0.02697733,  0.13758059, -0.00723211, -0.07190371,\n",
       "            0.00374287,  0.1102009 , -0.14125925, -0.05581926,\n",
       "           -0.02409938, -0.11838917,  0.10546297, -0.08293615,\n",
       "           -0.13804118,  0.04925409,  0.02976646,  0.03756236,\n",
       "           -0.08818024, -0.02973054,  0.04106247, -0.0329349 ,\n",
       "            0.01800823, -0.04657928, -0.12957637,  0.11254327]]],\n",
       " \n",
       " \n",
       "        [[[ 0.03579855, -0.11285061,  0.08453234, -0.10232786,\n",
       "            0.00059235,  0.12346683, -0.02969078, -0.10920084,\n",
       "           -0.13021788,  0.12425594, -0.01722177,  0.11690484,\n",
       "           -0.0066486 , -0.08884304,  0.01479405, -0.01593934,\n",
       "            0.13683917, -0.11688839, -0.1201755 , -0.07973067,\n",
       "            0.12499247,  0.06518428,  0.07829234, -0.08787064,\n",
       "           -0.13147789,  0.01627043,  0.05430833, -0.08930634,\n",
       "            0.11065187, -0.00802407,  0.06328663,  0.07082331]],\n",
       " \n",
       "         [[ 0.0491351 ,  0.05812286, -0.10942747,  0.02982584,\n",
       "           -0.10678133, -0.10616346,  0.12180333,  0.06526314,\n",
       "           -0.11486191, -0.02356356,  0.05178891,  0.13079996,\n",
       "           -0.14175028,  0.08821635,  0.00644599, -0.05946983,\n",
       "           -0.07813857, -0.09260748,  0.0098864 ,  0.12183036,\n",
       "            0.13484506, -0.05701764,  0.13709114, -0.09324437,\n",
       "           -0.12438665,  0.05505338,  0.03660862, -0.04078802,\n",
       "           -0.09343398,  0.14049338, -0.07238562, -0.02733412]],\n",
       " \n",
       "         [[ 0.1137331 ,  0.06802166, -0.03586421,  0.10895933,\n",
       "           -0.02198879, -0.10775571, -0.12320051, -0.03572392,\n",
       "           -0.11375444, -0.06626261, -0.01746144, -0.00839499,\n",
       "            0.03103878,  0.05816634,  0.09696172,  0.08503152,\n",
       "           -0.12977368,  0.04244785, -0.0458632 ,  0.0398659 ,\n",
       "            0.12249826, -0.09471332, -0.05494613,  0.11624958,\n",
       "            0.09225082, -0.12976813, -0.00937617,  0.13702635,\n",
       "            0.06451768, -0.02622056,  0.04830278,  0.03886357]]]],\n",
       "       dtype=float32),\n",
       " array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       dtype=float32),\n",
       " array([[[[ 0.04954401, -0.08249744,  0.00928581, ..., -0.04256659,\n",
       "            0.02179515,  0.06868065],\n",
       "          [ 0.05465824,  0.08284251,  0.00461831, ...,  0.07264004,\n",
       "            0.01616039,  0.06993634],\n",
       "          [-0.00802624, -0.01283152, -0.07706035, ..., -0.01131199,\n",
       "            0.0282943 ,  0.03271011],\n",
       "          ...,\n",
       "          [-0.04183953, -0.01899286, -0.02693228, ...,  0.00931048,\n",
       "            0.05522022, -0.02062901],\n",
       "          [ 0.015558  , -0.04449276,  0.02680081, ...,  0.03575027,\n",
       "           -0.03749834, -0.06969646],\n",
       "          [ 0.043171  , -0.00749702, -0.05111138, ..., -0.06841233,\n",
       "           -0.06435777, -0.01855326]],\n",
       " \n",
       "         [[-0.03555238,  0.02308316,  0.03744739, ...,  0.02005175,\n",
       "            0.03863621,  0.03476179],\n",
       "          [-0.00653285, -0.05570189, -0.07402924, ...,  0.00708669,\n",
       "            0.02779429,  0.05143588],\n",
       "          [ 0.01604956,  0.01775716, -0.04863564, ...,  0.07284857,\n",
       "            0.01537406, -0.00977166],\n",
       "          ...,\n",
       "          [-0.0014706 ,  0.0651025 , -0.0666717 , ..., -0.00603358,\n",
       "           -0.06283838, -0.00353578],\n",
       "          [ 0.02972112, -0.0085854 , -0.06617169, ..., -0.03550635,\n",
       "           -0.05414111, -0.06295101],\n",
       "          [ 0.02566288,  0.06174091, -0.07205335, ..., -0.08056778,\n",
       "            0.05469752,  0.03305888]],\n",
       " \n",
       "         [[ 0.07485519, -0.01817808,  0.05427188, ...,  0.05614994,\n",
       "            0.0191837 ,  0.02142161],\n",
       "          [ 0.02379772,  0.05536253, -0.08033727, ...,  0.009228  ,\n",
       "            0.01664609, -0.025803  ],\n",
       "          [ 0.03431601, -0.04861632, -0.08041053, ...,  0.01610405,\n",
       "            0.06372298,  0.07043416],\n",
       "          ...,\n",
       "          [ 0.0038391 ,  0.00585546,  0.07086349, ..., -0.01594145,\n",
       "           -0.01039949, -0.0700009 ],\n",
       "          [-0.05291637, -0.04420268, -0.05606926, ...,  0.07493096,\n",
       "           -0.05240868,  0.00467577],\n",
       "          [-0.06744291, -0.00438746,  0.00289932, ..., -0.0763031 ,\n",
       "            0.00206661, -0.03484694]]],\n",
       " \n",
       " \n",
       "        [[[-0.07696749, -0.07827969,  0.06255165, ...,  0.0734758 ,\n",
       "           -0.06112043, -0.07104167],\n",
       "          [-0.04691822,  0.01094859,  0.0411533 , ..., -0.07277252,\n",
       "            0.00365651,  0.02969537],\n",
       "          [-0.05243003, -0.03705078,  0.05158163, ..., -0.06109278,\n",
       "            0.03808149, -0.06717075],\n",
       "          ...,\n",
       "          [-0.03261794, -0.05577771,  0.01160982, ...,  0.05854521,\n",
       "            0.07262289,  0.0580428 ],\n",
       "          [-0.04067279,  0.03063848, -0.05978579, ...,  0.01593823,\n",
       "            0.01371182,  0.02098468],\n",
       "          [-0.00445371, -0.02733885, -0.07862759, ...,  0.04525232,\n",
       "           -0.04307069, -0.03282874]],\n",
       " \n",
       "         [[-0.04566715,  0.07184383,  0.02089622, ...,  0.00983542,\n",
       "           -0.02622759, -0.01217163],\n",
       "          [ 0.0054813 , -0.02598926,  0.06143894, ...,  0.0194765 ,\n",
       "           -0.0139019 , -0.04844615],\n",
       "          [ 0.02875074, -0.04631294,  0.07237633, ...,  0.04897269,\n",
       "            0.04414982,  0.05706584],\n",
       "          ...,\n",
       "          [-0.01869737, -0.0426571 ,  0.0077385 , ...,  0.05606008,\n",
       "           -0.02787813,  0.00412884],\n",
       "          [-0.07613697,  0.04255205,  0.07624754, ..., -0.0446719 ,\n",
       "            0.08150492,  0.02557653],\n",
       "          [ 0.07904191, -0.03264485, -0.02735916, ..., -0.08178837,\n",
       "            0.03126386,  0.04522105]],\n",
       " \n",
       "         [[ 0.04899838,  0.05340565,  0.0312608 , ...,  0.05329797,\n",
       "           -0.03878238,  0.06153352],\n",
       "          [-0.05250621,  0.05035972, -0.01360261, ..., -0.01499548,\n",
       "            0.069639  ,  0.02564178],\n",
       "          [-0.07899547, -0.06503119, -0.0042465 , ..., -0.04210529,\n",
       "           -0.07293835, -0.05599917],\n",
       "          ...,\n",
       "          [-0.07643861,  0.01997123,  0.04083568, ..., -0.04990585,\n",
       "           -0.00937369,  0.02999488],\n",
       "          [-0.06380835,  0.07250626,  0.07004728, ...,  0.0473055 ,\n",
       "            0.02274337, -0.03116722],\n",
       "          [ 0.03312021, -0.05348102,  0.06266191, ...,  0.04254941,\n",
       "            0.08070227,  0.02937853]]],\n",
       " \n",
       " \n",
       "        [[[ 0.03920823,  0.07658891,  0.00019803, ..., -0.04351578,\n",
       "           -0.06053514,  0.01768091],\n",
       "          [ 0.00081914,  0.04228615, -0.01889033, ..., -0.01382739,\n",
       "           -0.01286481, -0.02193196],\n",
       "          [-0.02300594, -0.02420698, -0.04110383, ..., -0.03663471,\n",
       "           -0.0417692 , -0.03498664],\n",
       "          ...,\n",
       "          [-0.00931916, -0.044878  ,  0.03543216, ..., -0.00261433,\n",
       "            0.00734797, -0.01590321],\n",
       "          [ 0.04865623, -0.01313344, -0.01922219, ...,  0.07148664,\n",
       "            0.05563477, -0.05502118],\n",
       "          [-0.04042246, -0.0398442 , -0.07359226, ..., -0.00392773,\n",
       "            0.0107178 , -0.01097059]],\n",
       " \n",
       "         [[-0.0011941 ,  0.04589307,  0.00557318, ...,  0.00432822,\n",
       "           -0.00147228,  0.02090368],\n",
       "          [-0.00984126,  0.02880722, -0.02583365, ...,  0.03017695,\n",
       "            0.05755296,  0.01559483],\n",
       "          [ 0.07035416, -0.01447266,  0.03621951, ..., -0.06241407,\n",
       "            0.00456373,  0.04233436],\n",
       "          ...,\n",
       "          [-0.00091437,  0.04329628,  0.06608846, ..., -0.05079767,\n",
       "           -0.06695237,  0.05834187],\n",
       "          [-0.08157207,  0.07088713,  0.00128029, ..., -0.06382662,\n",
       "            0.02699258, -0.0244202 ],\n",
       "          [ 0.06524468,  0.01168602, -0.05075923, ..., -0.02004807,\n",
       "            0.06038488, -0.08173792]],\n",
       " \n",
       "         [[ 0.07657067, -0.01215861,  0.03917906, ..., -0.01305061,\n",
       "           -0.03314319, -0.02609421],\n",
       "          [ 0.05971802,  0.07132313, -0.03637498, ...,  0.06145907,\n",
       "           -0.06563534,  0.08058985],\n",
       "          [ 0.013616  , -0.07938883,  0.07535917, ...,  0.06684179,\n",
       "           -0.07759875,  0.05301813],\n",
       "          ...,\n",
       "          [-0.04382371,  0.05019585,  0.01672989, ..., -0.04132665,\n",
       "            0.00448659, -0.05588277],\n",
       "          [ 0.01809625, -0.00146977,  0.05133059, ...,  0.08094303,\n",
       "            0.0478302 ,  0.0493587 ],\n",
       "          [-0.00297152,  0.0509946 , -0.07983173, ...,  0.01632611,\n",
       "            0.03708444, -0.07420354]]]], dtype=float32),\n",
       " array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.], dtype=float32),\n",
       " array([[[[ 0.01437951, -0.04525401, -0.0298059 , ...,  0.01092638,\n",
       "            0.02830605, -0.02525938],\n",
       "          [ 0.02034145,  0.05261293,  0.03899839, ...,  0.0522124 ,\n",
       "            0.03434562, -0.04409359],\n",
       "          [ 0.0478693 , -0.01962609,  0.03062271, ..., -0.00251991,\n",
       "            0.02132187, -0.02066358],\n",
       "          ...,\n",
       "          [-0.03690154,  0.04492432,  0.02436117, ..., -0.04253096,\n",
       "            0.01950949,  0.00665072],\n",
       "          [ 0.03647072,  0.05101523,  0.02971955, ...,  0.0281286 ,\n",
       "           -0.03267075,  0.01763195],\n",
       "          [ 0.05315256, -0.00063463,  0.047806  , ..., -0.02874169,\n",
       "            0.01344298,  0.00193203]],\n",
       " \n",
       "         [[ 0.03358043,  0.05140955,  0.02809643, ...,  0.04569675,\n",
       "           -0.04125962,  0.02470677],\n",
       "          [-0.04988264, -0.04231526, -0.01502889, ..., -0.02440792,\n",
       "            0.01214445, -0.05804944],\n",
       "          [ 0.00513472, -0.028335  , -0.04001584, ...,  0.04793103,\n",
       "           -0.0115797 , -0.03221395],\n",
       "          ...,\n",
       "          [ 0.0032416 , -0.00877472,  0.02842289, ..., -0.04013704,\n",
       "            0.03891047,  0.03700588],\n",
       "          [-0.04520576,  0.03825991, -0.05719246, ...,  0.05433044,\n",
       "            0.00708772, -0.00892808],\n",
       "          [ 0.04487738, -0.05105444,  0.02355387, ..., -0.01955391,\n",
       "            0.05532918,  0.02845889]],\n",
       " \n",
       "         [[ 0.0042106 , -0.03303618, -0.03387321, ...,  0.03418798,\n",
       "           -0.00950913,  0.0242119 ],\n",
       "          [ 0.0114556 ,  0.03586316, -0.05377021, ...,  0.00120619,\n",
       "            0.02181862, -0.03589883],\n",
       "          [ 0.03970948, -0.00666968, -0.01255237, ..., -0.0146644 ,\n",
       "           -0.00481094,  0.04488674],\n",
       "          ...,\n",
       "          [ 0.03845326, -0.02730808,  0.01006708, ...,  0.04091882,\n",
       "           -0.04566316, -0.02865742],\n",
       "          [-0.03257863,  0.00070614,  0.03806144, ..., -0.05811044,\n",
       "            0.01650385,  0.05330151],\n",
       "          [-0.03030524, -0.0166361 , -0.00925598, ..., -0.0306728 ,\n",
       "            0.04296647,  0.00794898]]],\n",
       " \n",
       " \n",
       "        [[[ 0.01393722, -0.00603943, -0.04307842, ..., -0.05071741,\n",
       "            0.00027425, -0.02301889],\n",
       "          [-0.04825315, -0.01911368,  0.01482195, ..., -0.0408062 ,\n",
       "           -0.01020245,  0.00697667],\n",
       "          [-0.0116512 ,  0.01362374,  0.05107801, ..., -0.04329224,\n",
       "           -0.01989496,  0.03722586],\n",
       "          ...,\n",
       "          [-0.00822683,  0.03690778, -0.05857122, ..., -0.05313089,\n",
       "            0.00208569, -0.00893923],\n",
       "          [ 0.01364471, -0.04375013, -0.05594254, ...,  0.0476262 ,\n",
       "           -0.04836202,  0.03960269],\n",
       "          [ 0.0342483 , -0.00268637, -0.0221942 , ..., -0.04999483,\n",
       "           -0.00078861,  0.00025944]],\n",
       " \n",
       "         [[ 0.00866309,  0.03652963,  0.03696479, ...,  0.05686005,\n",
       "           -0.03772296,  0.02243169],\n",
       "          [-0.04952993,  0.01679691,  0.00635643, ...,  0.04379097,\n",
       "            0.01179231, -0.01623109],\n",
       "          [-0.04573898,  0.03330756,  0.04245435, ..., -0.05157868,\n",
       "           -0.02340902,  0.03596025],\n",
       "          ...,\n",
       "          [ 0.01918396, -0.05484655, -0.04839491, ...,  0.01347208,\n",
       "            0.01080958, -0.02920432],\n",
       "          [ 0.05798603, -0.01996804,  0.01193452, ..., -0.02735703,\n",
       "           -0.04643597,  0.04764603],\n",
       "          [ 0.0441282 ,  0.01340519, -0.00070002, ..., -0.02666072,\n",
       "           -0.00624842, -0.0507225 ]],\n",
       " \n",
       "         [[-0.02205727, -0.01788046, -0.05866099, ...,  0.04012476,\n",
       "           -0.05825556,  0.04273628],\n",
       "          [-0.04185461,  0.00946082, -0.02573504, ...,  0.05047758,\n",
       "            0.00932204,  0.02939161],\n",
       "          [-0.01666235,  0.01141204, -0.05386755, ..., -0.00136472,\n",
       "            0.00868737,  0.00064265],\n",
       "          ...,\n",
       "          [-0.01051237,  0.0223967 , -0.0208926 , ...,  0.0309428 ,\n",
       "           -0.05648355,  0.05701244],\n",
       "          [ 0.03316118, -0.04214527,  0.02892317, ...,  0.01934837,\n",
       "           -0.05600884,  0.01373215],\n",
       "          [-0.02326116, -0.01993075,  0.04822405, ..., -0.04875148,\n",
       "           -0.01444762, -0.04180558]]],\n",
       " \n",
       " \n",
       "        [[[ 0.00729262, -0.01294781,  0.02937712, ...,  0.03702717,\n",
       "           -0.01981093,  0.00166668],\n",
       "          [ 0.04991917, -0.02126348,  0.05291923, ..., -0.01000966,\n",
       "            0.02587947, -0.0455166 ],\n",
       "          [ 0.01606726,  0.01295676, -0.00822377, ...,  0.04609438,\n",
       "           -0.00520388, -0.05152521],\n",
       "          ...,\n",
       "          [-0.02245252,  0.03977138,  0.03602198, ..., -0.03449863,\n",
       "            0.01857785,  0.04257361],\n",
       "          [-0.05113205, -0.00635243, -0.01353293, ..., -0.05674382,\n",
       "           -0.03006374,  0.03941966],\n",
       "          [-0.05863282, -0.01479769, -0.01600023, ..., -0.01985928,\n",
       "           -0.03625296,  0.01804366]],\n",
       " \n",
       "         [[ 0.01590505, -0.05689283, -0.0194496 , ...,  0.05318279,\n",
       "           -0.01825027,  0.04566194],\n",
       "          [-0.03565916,  0.03291002,  0.0177496 , ...,  0.00766129,\n",
       "            0.02089911, -0.05281001],\n",
       "          [ 0.05307051,  0.03199258,  0.03932241, ..., -0.05807685,\n",
       "           -0.03554548, -0.05180317],\n",
       "          ...,\n",
       "          [ 0.02632098, -0.02361259, -0.02840658, ...,  0.00283963,\n",
       "            0.05607656,  0.00983867],\n",
       "          [-0.04687551,  0.01221777, -0.05579289, ..., -0.01847055,\n",
       "            0.05519575, -0.00682775],\n",
       "          [ 0.03985072,  0.00242689,  0.02852197, ..., -0.02589844,\n",
       "           -0.03708327,  0.02068016]],\n",
       " \n",
       "         [[-0.05475101,  0.02180928, -0.0387393 , ..., -0.00203488,\n",
       "           -0.04745188, -0.0053643 ],\n",
       "          [-0.03404571, -0.04972295,  0.03472228, ...,  0.05739616,\n",
       "           -0.05349252, -0.0544775 ],\n",
       "          [-0.04609074, -0.05147814,  0.04842249, ...,  0.02298609,\n",
       "           -0.04652845,  0.02730027],\n",
       "          ...,\n",
       "          [-0.00584091, -0.02443621,  0.00615042, ...,  0.0076839 ,\n",
       "            0.01470752,  0.02025678],\n",
       "          [ 0.05440193,  0.0102005 ,  0.03750965, ..., -0.02686738,\n",
       "           -0.02977314, -0.031338  ],\n",
       "          [-0.02574364,  0.02956017, -0.05554676, ...,  0.0258753 ,\n",
       "           -0.00286783,  0.01697146]]]], dtype=float32),\n",
       " array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0.], dtype=float32),\n",
       " array([[-0.03293492,  0.03731779,  0.01190604, ..., -0.04185057,\n",
       "          0.02987177,  0.00402585],\n",
       "        [-0.01081318, -0.03278883, -0.01300256, ...,  0.01963356,\n",
       "         -0.03251185,  0.03587627],\n",
       "        [-0.03357008, -0.05162136, -0.03101698, ..., -0.05108765,\n",
       "         -0.02087739, -0.00311375],\n",
       "        ...,\n",
       "        [ 0.00054264, -0.03275298,  0.04912246, ...,  0.01163967,\n",
       "          0.00772496,  0.03128775],\n",
       "        [-0.0158951 , -0.0389621 ,  0.03650017, ..., -0.05119949,\n",
       "         -0.04950717,  0.03125383],\n",
       "        [ 0.03537591, -0.0346869 ,  0.03373082, ...,  0.05087205,\n",
       "         -0.00524117,  0.00090476]], dtype=float32),\n",
       " array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0.], dtype=float32),\n",
       " array([[-0.07176894,  0.04993813,  0.03164345, ...,  0.18220748,\n",
       "          0.00664088, -0.00181679],\n",
       "        [-0.00256066, -0.00238599, -0.08734618, ..., -0.0171812 ,\n",
       "         -0.03931847,  0.0787221 ],\n",
       "        [ 0.04753242,  0.17916413, -0.1286299 , ...,  0.1600023 ,\n",
       "          0.17720328, -0.01436117],\n",
       "        ...,\n",
       "        [ 0.14997764, -0.04723921, -0.11322422, ...,  0.00477669,\n",
       "          0.05158313,  0.07932065],\n",
       "        [ 0.00611013, -0.01890557, -0.00687155, ..., -0.16725798,\n",
       "          0.09723188, -0.15509221],\n",
       "        [ 0.19716819, -0.1721927 , -0.02634601, ..., -0.12614861,\n",
       "         -0.16479035, -0.14738505]], dtype=float32),\n",
       " array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0.], dtype=float32)]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fashion_model.get_weights()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_1 (Conv2D)            (None, 28, 28, 32)        320       \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_1 (LeakyReLU)    (None, 28, 28, 32)        0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 14, 14, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 14, 14, 64)        18496     \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_2 (LeakyReLU)    (None, 14, 14, 64)        0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 7, 7, 64)          0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 7, 7, 128)         73856     \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_3 (LeakyReLU)    (None, 7, 7, 128)         0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 4, 4, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 2048)              0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 128)               262272    \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_4 (LeakyReLU)    (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 10)                1290      \n",
      "=================================================================\n",
      "Total params: 356,234\n",
      "Trainable params: 356,234\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "fashion_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 48000 samples, validate on 12000 samples\n",
      "Epoch 1/2\n",
      "48000/48000 [==============================] - 206s 4ms/step - loss: 0.4703 - acc: 0.8297 - val_loss: 0.3449 - val_acc: 0.8766\n",
      "Epoch 2/2\n",
      "48000/48000 [==============================] - 204s 4ms/step - loss: 0.2963 - acc: 0.8909 - val_loss: 0.3134 - val_acc: 0.8858\n"
     ]
    }
   ],
   "source": [
    "fashion_train = fashion_model.fit(train_X, train_label, batch_size=batch_size,epochs =epochs, verbose=1, validation_data=(valid_X,valid_label))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_eval = fashion_model.evaluate(test_X, test_Y_one_hot, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test loss: 0.32280341728925704\n",
      "Test accuracy: 0.8783\n"
     ]
    }
   ],
   "source": [
    "print('Test loss:', test_eval[0])\n",
    "print('Test accuracy:', test_eval[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f35cd7cd30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f35cd9c7f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "accuracy= fashion_train.history['acc']\n",
    "val_accuracy = fashion_train.history['val_acc']\n",
    "loss = fashion_train.history['loss']\n",
    "val_loss = fashion_train.history['val_loss']\n",
    "epochs = range(len(accuracy))\n",
    "plt.plot(epochs, accuracy, 'bo', label='Training accuracy')\n",
    "plt.plot(epochs, val_accuracy, 'b', label='validation accuracy')\n",
    "plt.title('Training and validation accuracy')\n",
    "plt.legend()\n",
    "plt.figure()\n",
    "plt.plot(epochs, loss,'bo', label='Training loss')\n",
    "plt.plot(epochs, val_loss, 'b', label='validation loss')\n",
    "plt.title('Training and validation loss')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}