Fashion_MNIST_TF.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "universal-frederick",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Author -- Berkant Bayraktar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "fundamental-fitting",
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.keras import models\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras import optimizers\n",
    "from tensorflow.keras import Input\n",
    "from tensorflow.keras import utils\n",
    "from tensorflow.keras import datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "partial-granny",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "close-stations",
   "metadata": {},
   "outputs": [],
   "source": [
    "import requests\n",
    "requests.packages.urllib3.disable_warnings()\n",
    "import ssl\n",
    "\n",
    "try:\n",
    "    _create_unverified_https_context = ssl._create_unverified_context\n",
    "except AttributeError:\n",
    "    # Legacy Python that doesn't verify HTTPS certificates by default\n",
    "    pass\n",
    "else:\n",
    "    # Handle target environment that doesn't support HTTPS verification\n",
    "    ssl._create_default_https_context = _create_unverified_https_context"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "unexpected-butler",
   "metadata": {},
   "outputs": [],
   "source": [
    "fashion_mnist = datasets.fashion_mnist.load_data()    # Load FashionMNIST Dataset using Keras API"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "given-husband",
   "metadata": {},
   "outputs": [],
   "source": [
    "((x_train, y_train), (x_test, y_test)) = fashion_mnist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "sexual-vanilla",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train: X=(60000, 28, 28), y=(60000,)\n",
      "Test: X=(10000, 28, 28), y=(10000,)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print('Train: X=%s, y=%s' % (x_train.shape, y_train.shape))\n",
    "print('Test: X=%s, y=%s' % (x_test.shape, y_test.shape))\n",
    "# plot first few images\n",
    "for i in range(9):\n",
    "\t# define subplot\n",
    "\tplt.subplot(330 + 1 + i )\n",
    "\t# plot raw pixel data\n",
    "\tplt.imshow(x_train[i], cmap=plt.get_cmap('gray'))\n",
    "# show the figure\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "floral-treasure",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of classes :  10\n"
     ]
    }
   ],
   "source": [
    "print('Number of classes : ' , len(set(y_train)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "billion-arena",
   "metadata": {},
   "outputs": [],
   "source": [
    "(x_train, x_test) = (x_train / 255, x_test / 255)\n",
    "\n",
    "# convert x_train and y_train from (count, 28, 28) -> (count, 28, 28, 1)\n",
    "x_train = x_train[...,np.newaxis]\n",
    "x_test = x_test[...,np.newaxis]\n",
    "y_train = utils.to_categorical(y_train, 10)\n",
    "y_test = utils.to_categorical(y_test, 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "informed-assist",
   "metadata": {},
   "outputs": [],
   "source": [
    "# initialize the label names\n",
    "labelNames = [\"top\", \"trouser\", \"pullover\", \"dress\", \"coat\",\n",
    "\t\"sandal\", \"shirt\", \"sneaker\", \"bag\", \"ankle boot\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "heard-rotation",
   "metadata": {},
   "outputs": [],
   "source": [
    "class berkantNET:\n",
    "    def __init__(self, number_of_classes):\n",
    "        self.model = None\n",
    "        self.number_of_classes = number_of_classes\n",
    "    def createModel(self):\n",
    "        self.model = models.Sequential([\n",
    "            Input(shape = (28,28,1)), \n",
    "            layers.Conv2D(32,(3,3), strides=(1,1), padding= 'same', activation = 'relu', use_bias=True, name = \"layer_1\"),\n",
    "            layers.MaxPooling2D(pool_size = (2,2), strides = 2),\n",
    "            layers.Conv2D(64,(3,3), strides=(1,1), padding= 'same', activation = 'relu', use_bias=True, name = \"layer_2\"),\n",
    "            layers.MaxPooling2D(pool_size = (2,2), strides = 2),\n",
    "            layers.Conv2D(128,(3,3), strides=(1,1), padding= 'same', activation = 'relu', use_bias=True, name = \"layer_3\"),\n",
    "            layers.MaxPooling2D(pool_size = (2,2), strides = 2),\n",
    "            layers.Flatten(),\n",
    "            layers.Dropout(rate = 0.5),\n",
    "            layers.Dense(512, activation= 'relu'),\n",
    "            layers.Dropout(rate = 0.2),\n",
    "            layers.Dense(10, activation='softmax')\n",
    "        ])\n",
    "    def run(self, i):\n",
    "        print(self.model.predict(i))\n",
    "        \n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "integrated-receptor",
   "metadata": {},
   "outputs": [],
   "source": [
    "bnet = berkantNET(number_of_classes=10)\n",
    "bnet.createModel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "brilliant-median",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 60000 samples, validate on 10000 samples\n",
      "Epoch 1/10\n",
      "60000/60000 [==============================] - 13s 216us/sample - loss: 0.4925 - accuracy: 0.8179 - val_loss: 0.3322 - val_accuracy: 0.8735\n",
      "Epoch 2/10\n",
      "60000/60000 [==============================] - 12s 195us/sample - loss: 0.3203 - accuracy: 0.8813 - val_loss: 0.2869 - val_accuracy: 0.8944\n",
      "Epoch 3/10\n",
      "60000/60000 [==============================] - 12s 193us/sample - loss: 0.2788 - accuracy: 0.8977 - val_loss: 0.2494 - val_accuracy: 0.9083\n",
      "Epoch 4/10\n",
      "60000/60000 [==============================] - 12s 197us/sample - loss: 0.2525 - accuracy: 0.9061 - val_loss: 0.2477 - val_accuracy: 0.9081\n",
      "Epoch 5/10\n",
      "60000/60000 [==============================] - 12s 196us/sample - loss: 0.2349 - accuracy: 0.9126 - val_loss: 0.2387 - val_accuracy: 0.9118\n",
      "Epoch 6/10\n",
      "60000/60000 [==============================] - 12s 193us/sample - loss: 0.2163 - accuracy: 0.9194 - val_loss: 0.2270 - val_accuracy: 0.9200\n",
      "Epoch 7/10\n",
      "60000/60000 [==============================] - 11s 190us/sample - loss: 0.2072 - accuracy: 0.9226 - val_loss: 0.2196 - val_accuracy: 0.9204\n",
      "Epoch 8/10\n",
      "60000/60000 [==============================] - 12s 197us/sample - loss: 0.1986 - accuracy: 0.9262 - val_loss: 0.2198 - val_accuracy: 0.9196\n",
      "Epoch 9/10\n",
      "60000/60000 [==============================] - 12s 194us/sample - loss: 0.1896 - accuracy: 0.9300 - val_loss: 0.2166 - val_accuracy: 0.9207\n",
      "Epoch 10/10\n",
      "60000/60000 [==============================] - 12s 196us/sample - loss: 0.1790 - accuracy: 0.9333 - val_loss: 0.2208 - val_accuracy: 0.9196\n"
     ]
    }
   ],
   "source": [
    "optimizer = optimizers.Adam(learning_rate = 0.001)\n",
    "bnet.model.compile(optimizer = optimizer, loss = 'categorical_crossentropy', metrics= ['accuracy'])\n",
    "res = bnet.model.fit(x = x_train, y = y_train, epochs=10, batch_size = 32, validation_data = (x_test,y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "experienced-irrigation",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f2de89ef390>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(res.history['loss'], label = 'loss')\n",
    "plt.plot(res.history['val_loss'], label = 'val_loss')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "colored-efficiency",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f2dbc39fed0>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(res.history['accuracy'], label = 'accuracy')\n",
    "plt.plot(res.history['val_accuracy'], label = 'val_accuracy')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "japanese-resolution",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.99999857,\n",
       " 0.9996562,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.82147443,\n",
       " 1.0,\n",
       " 0.9948001,\n",
       " 0.99312395,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9975389,\n",
       " 0.9999682,\n",
       " 0.9661656,\n",
       " 0.9992778,\n",
       " 0.52756417,\n",
       " 1.0,\n",
       " 0.9591806,\n",
       " 0.6540904,\n",
       " 0.9999758,\n",
       " 0.99613416,\n",
       " 0.99547005,\n",
       " 0.96783197,\n",
       " 0.99999094,\n",
       " 0.97037965,\n",
       " 1.0,\n",
       " 0.83284765,\n",
       " 0.5670172,\n",
       " 0.80690885,\n",
       " 0.9999374,\n",
       " 0.6218027,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.98051274,\n",
       " 0.82539093,\n",
       " 1.0,\n",
       " 0.98741865,\n",
       " 1.0,\n",
       " 0.9995604,\n",
       " 0.999948,\n",
       " 1.0,\n",
       " 0.80579525,\n",
       " 1.0,\n",
       " 0.6292542,\n",
       " 0.7715732,\n",
       " 0.9121382,\n",
       " 0.8299852,\n",
       " 0.8459856,\n",
       " 1.0,\n",
       " 0.91305786,\n",
       " 0.6408809,\n",
       " 0.89390355,\n",
       " 0.6470006,\n",
       " 1.0,\n",
       " 0.99967253,\n",
       " 0.9995196,\n",
       " 0.96424216,\n",
       " 1.0,\n",
       " 0.6366651,\n",
       " 0.9993036,\n",
       " 0.99999285,\n",
       " 0.99986935,\n",
       " 0.99991393,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.4763891,\n",
       " 0.87354165,\n",
       " 0.9313497,\n",
       " 1.0,\n",
       " 0.9999831,\n",
       " 0.97232497,\n",
       " 0.9984079,\n",
       " 0.67305046,\n",
       " 0.97995913,\n",
       " 0.9764165,\n",
       " 1.0,\n",
       " 0.9993279,\n",
       " 0.99998915,\n",
       " 0.9986442,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9999999,\n",
       " 1.0,\n",
       " 0.9997954,\n",
       " 0.9885068,\n",
       " 0.9999932,\n",
       " 0.99999106,\n",
       " 0.6102756,\n",
       " 0.99999297,\n",
       " 0.9814507,\n",
       " 0.9922625,\n",
       " 0.9999988,\n",
       " 1.0,\n",
       " 0.9999554,\n",
       " 0.9682836,\n",
       " 1.0,\n",
       " 0.8098895,\n",
       " 0.9733565,\n",
       " 0.99902356,\n",
       " 0.9934702,\n",
       " 0.9999893,\n",
       " 0.7411218,\n",
       " 0.9999993,\n",
       " 0.9997793,\n",
       " 1.0,\n",
       " 0.71101457,\n",
       " 0.99999964,\n",
       " 0.995771,\n",
       " 0.99994767,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.80493325,\n",
       " 1.0,\n",
       " 0.99981934,\n",
       " 0.99999833,\n",
       " 0.8044084,\n",
       " 0.99925095,\n",
       " 0.99999726,\n",
       " 0.98702735,\n",
       " 0.999977,\n",
       " 0.981619,\n",
       " 0.99999857,\n",
       " 0.99315095,\n",
       " 0.99965036,\n",
       " 1.0,\n",
       " 0.880798,\n",
       " 1.0,\n",
       " 0.9999976,\n",
       " 0.99983025,\n",
       " 1.0,\n",
       " 0.9999994,\n",
       " 1.0,\n",
       " 0.99997914,\n",
       " 0.89883417,\n",
       " 0.9564572,\n",
       " 1.0,\n",
       " 0.9999815,\n",
       " 0.9999975,\n",
       " 1.0,\n",
       " 0.6941851,\n",
       " 0.9997781,\n",
       " 0.99965036,\n",
       " 1.0,\n",
       " 0.8256307,\n",
       " 1.0,\n",
       " 0.72766465,\n",
       " 0.99976605,\n",
       " 0.941314,\n",
       " 0.95202506,\n",
       " 0.97248703,\n",
       " 1.0,\n",
       " 0.5223507,\n",
       " 1.0,\n",
       " 0.5088564,\n",
       " 0.9517523,\n",
       " 0.99999964,\n",
       " 1.0,\n",
       " 0.9864336,\n",
       " 0.99864703,\n",
       " 0.9997551,\n",
       " 0.92529315,\n",
       " 0.999699,\n",
       " 0.9999715,\n",
       " 1.0,\n",
       " 0.7899381,\n",
       " 0.95205605,\n",
       " 0.99989545,\n",
       " 0.9865302,\n",
       " 0.7469298,\n",
       " 1.0,\n",
       " 0.9018073,\n",
       " 1.0,\n",
       " 0.9967445,\n",
       " 1.0,\n",
       " 0.999605,\n",
       " 0.98748183,\n",
       " 0.99999917,\n",
       " 1.0,\n",
       " 0.9998361,\n",
       " 0.9999999,\n",
       " 0.9237785,\n",
       " 0.8496978,\n",
       " 0.9999994,\n",
       " 0.99999964,\n",
       " 0.999998,\n",
       " 0.9969523,\n",
       " 0.97919214,\n",
       " 0.99999785,\n",
       " 0.99999976,\n",
       " 0.99981326,\n",
       " 0.9971156,\n",
       " 0.8399048,\n",
       " 0.9933361,\n",
       " 0.9773677,\n",
       " 0.9944154,\n",
       " 0.9958007,\n",
       " 0.99759704,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.99938273,\n",
       " 0.9916435,\n",
       " 0.9999027,\n",
       " 0.99987495,\n",
       " 0.99195695,\n",
       " 0.98101395,\n",
       " 0.99998593,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9997141,\n",
       " 1.0,\n",
       " 0.99995303,\n",
       " 0.764476,\n",
       " 1.0,\n",
       " 0.99979883,\n",
       " 1.0,\n",
       " 0.98673624,\n",
       " 0.9999453,\n",
       " 0.9985476,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.98380774,\n",
       " 0.9009717,\n",
       " 1.0,\n",
       " 0.9999862,\n",
       " 0.9987509,\n",
       " 0.62604225,\n",
       " 0.99999964,\n",
       " 0.93060803,\n",
       " 0.999997,\n",
       " 0.99999964,\n",
       " 1.0,\n",
       " 0.91053116,\n",
       " 1.0,\n",
       " 0.99652463,\n",
       " 1.0,\n",
       " 0.9999994,\n",
       " 0.9992593,\n",
       " 0.5844676,\n",
       " 1.0,\n",
       " 0.88446116,\n",
       " 0.99972564,\n",
       " 0.97934353,\n",
       " 0.6067533,\n",
       " 0.9999999,\n",
       " 0.97981447,\n",
       " 0.4973859,\n",
       " 0.5187355,\n",
       " 0.9953696,\n",
       " 0.99686277,\n",
       " 0.99999976,\n",
       " 0.63006943,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.5785149,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9863051,\n",
       " 0.99273187,\n",
       " 0.99999106,\n",
       " 0.9535632,\n",
       " 0.9999602,\n",
       " 0.9772768,\n",
       " 0.99995124,\n",
       " 0.99575293,\n",
       " 0.48741087,\n",
       " 1.0,\n",
       " 0.9999924,\n",
       " 0.8282456,\n",
       " 1.0,\n",
       " 0.9215247,\n",
       " 0.9997453,\n",
       " 0.9420914,\n",
       " 0.9985526,\n",
       " 0.99999774,\n",
       " 0.99821067,\n",
       " 0.86112475,\n",
       " 1.0,\n",
       " 0.96819055,\n",
       " 0.99967957,\n",
       " 1.0,\n",
       " 0.94727606,\n",
       " 0.5974552,\n",
       " 0.9913935,\n",
       " 0.5111654,\n",
       " 0.54252,\n",
       " 0.99999905,\n",
       " 0.99999976,\n",
       " 0.8644818,\n",
       " 0.9999962,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.99503785,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.99999297,\n",
       " 0.99994624,\n",
       " 0.9996841,\n",
       " 0.9662226,\n",
       " 0.9582918,\n",
       " 0.99091387,\n",
       " 0.9999932,\n",
       " 0.9821374,\n",
       " 0.9944634,\n",
       " 0.99355996,\n",
       " 1.0,\n",
       " 0.9999995,\n",
       " 0.9042312,\n",
       " 1.0,\n",
       " 0.7716323,\n",
       " 0.61080647,\n",
       " 0.80413204,\n",
       " 0.9968829,\n",
       " 0.9925211,\n",
       " 0.84636134,\n",
       " 1.0,\n",
       " 0.9967738,\n",
       " 0.9997218,\n",
       " 0.9985002,\n",
       " 0.9208729,\n",
       " 0.88922936,\n",
       " 1.0,\n",
       " 0.65548,\n",
       " 0.37416348,\n",
       " 0.9999994,\n",
       " 0.9999765,\n",
       " 1.0,\n",
       " 0.9949856,\n",
       " 0.999881,\n",
       " 0.62952006,\n",
       " 0.8308937,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9976146,\n",
       " 0.9999995,\n",
       " 0.9924176,\n",
       " 0.9990606,\n",
       " 0.98345435,\n",
       " 0.9963257,\n",
       " 0.7988268,\n",
       " 0.866958,\n",
       " 1.0,\n",
       " 0.83902717,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9991534,\n",
       " 0.9999995,\n",
       " 0.99999964,\n",
       " 1.0,\n",
       " 0.9999838,\n",
       " 0.5320946,\n",
       " 0.9952035,\n",
       " 0.83124024,\n",
       " 0.9984043,\n",
       " 0.98234,\n",
       " 0.98925215,\n",
       " 0.99999964,\n",
       " 0.90990156,\n",
       " 1.0,\n",
       " 0.584083,\n",
       " 0.99613684,\n",
       " 0.86133385,\n",
       " 0.9996681,\n",
       " 0.9999764,\n",
       " 0.99846256,\n",
       " 0.99824595,\n",
       " 0.9769712,\n",
       " 0.99999917,\n",
       " 0.94969666,\n",
       " 0.99999964,\n",
       " 0.99999595,\n",
       " 0.9983082,\n",
       " 0.6919,\n",
       " 1.0,\n",
       " 0.9970583,\n",
       " 0.999992,\n",
       " 0.63519424,\n",
       " 0.8339098,\n",
       " 1.0,\n",
       " 0.93152666,\n",
       " 0.8600127,\n",
       " 0.9989925,\n",
       " 1.0,\n",
       " 0.9999994,\n",
       " 1.0,\n",
       " 0.8278537,\n",
       " 0.9984459,\n",
       " 0.55206627,\n",
       " 0.9585932,\n",
       " 0.99988174,\n",
       " 0.99229574,\n",
       " 1.0,\n",
       " 0.57309395,\n",
       " 0.7203846,\n",
       " 0.8528617,\n",
       " 0.9701505,\n",
       " 1.0,\n",
       " 0.97559935,\n",
       " 0.9994886,\n",
       " 0.9999993,\n",
       " 0.9998659,\n",
       " 0.9839522,\n",
       " 0.91689265,\n",
       " 0.98958254,\n",
       " 0.56305844,\n",
       " 0.58106947,\n",
       " 0.9468873,\n",
       " 0.99746203,\n",
       " 0.471748,\n",
       " 0.9909792,\n",
       " 0.99999475,\n",
       " 0.68584853,\n",
       " 0.9999999,\n",
       " 0.99878675,\n",
       " 0.9999268,\n",
       " 1.0,\n",
       " 0.9999969,\n",
       " 0.96967363,\n",
       " 1.0,\n",
       " 0.99901557,\n",
       " 0.9998803,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9362089,\n",
       " 0.99044096,\n",
       " 1.0,\n",
       " 0.9999994,\n",
       " 1.0,\n",
       " 0.9972077,\n",
       " 0.99448615,\n",
       " 0.9994392,\n",
       " 0.9979944,\n",
       " 0.99132043,\n",
       " 0.76294965,\n",
       " 1.0,\n",
       " 0.8657331,\n",
       " 1.0,\n",
       " 0.99182117,\n",
       " 0.9901883,\n",
       " 0.96733797,\n",
       " 1.0,\n",
       " 0.99999833,\n",
       " 0.7238586,\n",
       " 0.99996984,\n",
       " 0.97335815,\n",
       " 1.0,\n",
       " 0.9999926,\n",
       " 0.99998605,\n",
       " 0.9895823,\n",
       " 0.9276437,\n",
       " 0.5164603,\n",
       " 1.0,\n",
       " 0.5395439,\n",
       " 0.9320513,\n",
       " 1.0,\n",
       " 0.8189243,\n",
       " 0.99879766,\n",
       " 0.8698331,\n",
       " 0.55789596,\n",
       " 0.98792577,\n",
       " 0.99401146,\n",
       " 0.9997793,\n",
       " 0.99971825,\n",
       " 0.9969408,\n",
       " 0.9991641,\n",
       " 0.99999964,\n",
       " 0.9972886,\n",
       " 1.0,\n",
       " 0.9999269,\n",
       " 0.96376026,\n",
       " 0.9999993,\n",
       " 0.9437561,\n",
       " 0.51423335,\n",
       " 0.99877864,\n",
       " 0.5246389,\n",
       " 0.99935275,\n",
       " 0.88110495,\n",
       " 1.0,\n",
       " 0.99999976,\n",
       " 1.0,\n",
       " 0.9902374,\n",
       " 0.97795403,\n",
       " 1.0,\n",
       " 0.8453173,\n",
       " 0.99999976,\n",
       " 1.0,\n",
       " 0.9192325,\n",
       " 0.9999987,\n",
       " 0.55886066,\n",
       " 0.9016058,\n",
       " 1.0,\n",
       " 0.9999993,\n",
       " 0.99999833,\n",
       " 0.867862,\n",
       " 1.0,\n",
       " 0.9999665,\n",
       " 0.6798633,\n",
       " 0.89737725,\n",
       " 0.9949384,\n",
       " 0.99995756,\n",
       " 1.0,\n",
       " 0.979463,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.99502337,\n",
       " 0.9948419,\n",
       " 0.999632,\n",
       " 0.99989045,\n",
       " 0.9343791,\n",
       " 0.8283357,\n",
       " 0.8865605,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9965838,\n",
       " 1.0,\n",
       " 0.98764217,\n",
       " 0.9999994,\n",
       " 0.9999999,\n",
       " 0.88268256,\n",
       " 0.59260374,\n",
       " 0.94724035,\n",
       " 1.0,\n",
       " 0.6981073,\n",
       " 0.9969652,\n",
       " 0.7161095,\n",
       " 1.0,\n",
       " 0.9999999,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.6879663,\n",
       " 1.0,\n",
       " 0.9979563,\n",
       " 0.9986609,\n",
       " 0.9959275,\n",
       " 0.79601556,\n",
       " 0.99999976,\n",
       " 1.0,\n",
       " 0.99372417,\n",
       " 1.0,\n",
       " 0.99664366,\n",
       " 0.98124427,\n",
       " 0.7179334,\n",
       " 0.99995124,\n",
       " 0.9969388,\n",
       " 1.0,\n",
       " 0.99969363,\n",
       " 1.0,\n",
       " 0.99999905,\n",
       " 1.0,\n",
       " 0.99994195,\n",
       " 0.9841724,\n",
       " 0.6247013,\n",
       " 1.0,\n",
       " 0.9982667,\n",
       " 0.6622574,\n",
       " 0.9755117,\n",
       " 0.92963314,\n",
       " 0.9988199,\n",
       " 0.7817432,\n",
       " 0.9999745,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.67790127,\n",
       " 0.82207453,\n",
       " 0.9997826,\n",
       " 0.83256435,\n",
       " 0.61273426,\n",
       " 0.9999999,\n",
       " 0.99999964,\n",
       " 0.9999895,\n",
       " 0.6654683,\n",
       " 0.63562334,\n",
       " 0.99999976,\n",
       " 0.9999863,\n",
       " 0.95361435,\n",
       " 0.999967,\n",
       " 1.0,\n",
       " 0.9494328,\n",
       " 0.94100356,\n",
       " 0.61664444,\n",
       " 0.52306616,\n",
       " 1.0,\n",
       " 0.9816259,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.99999976,\n",
       " 0.7543284,\n",
       " 0.47538438,\n",
       " 0.9999999,\n",
       " 0.99999785,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.6883204,\n",
       " 1.0,\n",
       " 0.96952033,\n",
       " 0.99997544,\n",
       " 1.0,\n",
       " 0.99998677,\n",
       " 1.0,\n",
       " 0.72565013,\n",
       " 0.94981915,\n",
       " 0.9999976,\n",
       " 0.9939294,\n",
       " 1.0,\n",
       " 0.5239123,\n",
       " 0.9999527,\n",
       " 0.98211944,\n",
       " 1.0,\n",
       " 0.9351791,\n",
       " 0.99575883,\n",
       " 0.9904232,\n",
       " 0.9999999,\n",
       " 0.9992235,\n",
       " 0.96786165,\n",
       " 0.9499608,\n",
       " 0.98351014,\n",
       " 0.9422063,\n",
       " 0.99860674,\n",
       " 1.0,\n",
       " 0.99992585,\n",
       " 0.6483175,\n",
       " 0.8735884,\n",
       " 0.9988255,\n",
       " 0.99983716,\n",
       " 0.9941864,\n",
       " 0.99434966,\n",
       " 0.7671705,\n",
       " 0.71981454,\n",
       " 0.99826556,\n",
       " 0.9178126,\n",
       " 0.99945885,\n",
       " 0.86401975,\n",
       " 0.9998097,\n",
       " 0.9999999,\n",
       " 1.0,\n",
       " 0.99434894,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9801214,\n",
       " 0.9999995,\n",
       " 0.9999478,\n",
       " 0.9995968,\n",
       " 0.99762636,\n",
       " 0.80149585,\n",
       " 0.99946564,\n",
       " 0.9998369,\n",
       " 0.99999964,\n",
       " 0.965375,\n",
       " 0.9884291,\n",
       " 0.95642656,\n",
       " 0.99997914,\n",
       " 0.58395314,\n",
       " 1.0,\n",
       " 0.538549,\n",
       " 1.0,\n",
       " 0.99154776,\n",
       " 0.52791464,\n",
       " 0.86381006,\n",
       " 0.99781644,\n",
       " 0.9999895,\n",
       " 0.99892646,\n",
       " 0.6465366,\n",
       " 0.5119209,\n",
       " 0.9978543,\n",
       " 0.9967818,\n",
       " 0.9998167,\n",
       " 0.5137641,\n",
       " 0.9999987,\n",
       " 1.0,\n",
       " 0.99966645,\n",
       " 1.0,\n",
       " 0.99996865,\n",
       " 0.99999774,\n",
       " 0.45851746,\n",
       " 1.0,\n",
       " 0.9997392,\n",
       " 0.99442613,\n",
       " 0.83828616,\n",
       " 0.99262035,\n",
       " 0.99999774,\n",
       " 0.5107088,\n",
       " 0.4742568,\n",
       " 1.0,\n",
       " 0.9986846,\n",
       " 0.9999709,\n",
       " 1.0,\n",
       " 0.99998224,\n",
       " 0.9974324,\n",
       " 0.992611,\n",
       " 0.999995,\n",
       " 0.97546023,\n",
       " 0.9443624,\n",
       " 1.0,\n",
       " 0.9999896,\n",
       " 0.70515084,\n",
       " 0.7043024,\n",
       " 1.0,\n",
       " 0.7070027,\n",
       " 0.99833894,\n",
       " 0.99498355,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.8745015,\n",
       " 0.999897,\n",
       " 0.5942776,\n",
       " 0.9999355,\n",
       " 0.91343284,\n",
       " 0.5509879,\n",
       " 0.7489314,\n",
       " 1.0,\n",
       " 0.9737138,\n",
       " 0.99865615,\n",
       " 0.9999999,\n",
       " 0.88046044,\n",
       " 0.99838865,\n",
       " 0.92924565,\n",
       " 0.78707546,\n",
       " 0.74960434,\n",
       " 0.99967086,\n",
       " 1.0,\n",
       " 0.9999988,\n",
       " 1.0,\n",
       " 0.99796,\n",
       " 0.9800347,\n",
       " 0.99999976,\n",
       " 0.9997664,\n",
       " 0.99999976,\n",
       " 0.66057783,\n",
       " 0.8658073,\n",
       " 0.81970245,\n",
       " 0.9683856,\n",
       " 0.99984086,\n",
       " 0.9991978,\n",
       " 1.0,\n",
       " 0.9999871,\n",
       " 0.9999939,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.6711143,\n",
       " 1.0,\n",
       " 0.99999666,\n",
       " 0.9995307,\n",
       " 0.9528888,\n",
       " 1.0,\n",
       " 0.527078,\n",
       " 1.0,\n",
       " 0.9999976,\n",
       " 0.9999995,\n",
       " 1.0,\n",
       " 0.96128815,\n",
       " 1.0,\n",
       " 0.5100087,\n",
       " 0.8884177,\n",
       " 0.57088137,\n",
       " 0.99999905,\n",
       " 0.92661035,\n",
       " 0.7019086,\n",
       " 0.99939775,\n",
       " 0.98098433,\n",
       " 0.9999993,\n",
       " 0.9998621,\n",
       " 0.9966723,\n",
       " 1.0,\n",
       " 0.9999988,\n",
       " 0.9987092,\n",
       " 0.9994248,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9647724,\n",
       " 0.9995297,\n",
       " 0.9999629,\n",
       " 0.9910437,\n",
       " 0.99999595,\n",
       " 0.99699163,\n",
       " 1.0,\n",
       " 0.99947304,\n",
       " 0.99997556,\n",
       " 0.9952565,\n",
       " 0.91906965,\n",
       " 0.9999993,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.7904896,\n",
       " 0.9973514,\n",
       " 1.0,\n",
       " 0.9847083,\n",
       " 1.0,\n",
       " 0.5108057,\n",
       " 0.9624831,\n",
       " 0.72959274,\n",
       " 0.6131578,\n",
       " 0.99999976,\n",
       " 0.9620321,\n",
       " 0.997569,\n",
       " 0.51947993,\n",
       " 0.9999839,\n",
       " 0.99999976,\n",
       " 1.0,\n",
       " 0.997964,\n",
       " 1.0,\n",
       " 0.8738663,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.99599767,\n",
       " 0.97851264,\n",
       " 0.67733693,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.98304236,\n",
       " 0.99914265,\n",
       " 0.5492231,\n",
       " 0.9409458,\n",
       " 1.0,\n",
       " 0.91137904,\n",
       " 0.94884115,\n",
       " 0.9989318,\n",
       " 0.8897093,\n",
       " 0.5534159,\n",
       " 0.99999285,\n",
       " 1.0,\n",
       " 0.9994973,\n",
       " 0.90637964,\n",
       " 1.0,\n",
       " 0.993071,\n",
       " 0.9074101,\n",
       " 1.0,\n",
       " 0.9999999,\n",
       " 0.9999999,\n",
       " 0.9999827,\n",
       " 0.9999988,\n",
       " 1.0,\n",
       " 0.9667927,\n",
       " 1.0,\n",
       " 0.9999473,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.99999845,\n",
       " 0.99999654,\n",
       " 0.6115758,\n",
       " 0.9999999,\n",
       " 0.9999999,\n",
       " 1.0,\n",
       " 0.99999356,\n",
       " 0.9484381,\n",
       " 1.0,\n",
       " 0.9210658,\n",
       " 0.99663866,\n",
       " 0.5314746,\n",
       " 1.0,\n",
       " 0.9999993,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9999995,\n",
       " 0.9999995,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9784822,\n",
       " 0.6285496,\n",
       " 0.9829641,\n",
       " 0.9224851,\n",
       " 1.0,\n",
       " 0.9999999,\n",
       " 0.5410124,\n",
       " 1.0,\n",
       " 0.9053705,\n",
       " 0.9834907,\n",
       " 0.9999969,\n",
       " 1.0,\n",
       " 0.5645534,\n",
       " 0.9999999,\n",
       " 1.0,\n",
       " 0.99999976,\n",
       " 0.96685266,\n",
       " 1.0,\n",
       " 0.9947101,\n",
       " 0.993083,\n",
       " 0.99985206,\n",
       " 0.8566859,\n",
       " 1.0,\n",
       " 0.99999976,\n",
       " 0.9799992,\n",
       " 0.95947933,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9999832,\n",
       " 0.99997723,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.8570499,\n",
       " 0.9993538,\n",
       " 0.99890184,\n",
       " 0.7329302,\n",
       " 0.99830604,\n",
       " 0.9982268,\n",
       " 0.9773163,\n",
       " 0.7753969,\n",
       " 0.6305166,\n",
       " 0.99815756,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9993728,\n",
       " 0.8922614,\n",
       " 1.0,\n",
       " 0.8877526,\n",
       " 0.99999976,\n",
       " 0.98663783,\n",
       " 0.60533077,\n",
       " 0.99999976,\n",
       " 0.9994209,\n",
       " 0.99382234,\n",
       " 0.99945134,\n",
       " 0.921549,\n",
       " 0.9999994,\n",
       " 0.9874636,\n",
       " 0.99998057,\n",
       " 0.69426566,\n",
       " 0.6825096,\n",
       " 0.99953127,\n",
       " 0.9968681,\n",
       " 0.99990964,\n",
       " 0.94964033,\n",
       " 1.0,\n",
       " 0.699497,\n",
       " 0.7484272,\n",
       " 0.5493056,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.99999964,\n",
       " 1.0,\n",
       " 0.9894533,\n",
       " 0.999977,\n",
       " 0.9999807,\n",
       " 0.91548884,\n",
       " 0.60447264,\n",
       " 0.9997938,\n",
       " 1.0,\n",
       " 0.99535096,\n",
       " 1.0,\n",
       " 0.98777187,\n",
       " 0.9958591,\n",
       " 0.9980944,\n",
       " 0.42795876,\n",
       " 0.99999607,\n",
       " 0.9322302,\n",
       " 0.998862,\n",
       " 0.8304667,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.6677283,\n",
       " 1.0,\n",
       " 0.44326094,\n",
       " 0.86121863,\n",
       " 0.6194262,\n",
       " 0.6907303,\n",
       " 0.99844664,\n",
       " 0.99381155,\n",
       " 0.5515257,\n",
       " 0.992353,\n",
       " 0.9996371,\n",
       " 1.0,\n",
       " 0.99642855,\n",
       " 0.55460757,\n",
       " 0.6227827,\n",
       " 1.0,\n",
       " 0.9846731,\n",
       " 0.9999634,\n",
       " 0.9999831,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 0.9988121,\n",
       " 0.7854292,\n",
       " 0.9941462,\n",
       " 0.99510384,\n",
       " 0.99987507,\n",
       " 0.9989195,\n",
       " 0.99994254,\n",
       " 0.9986381,\n",
       " 0.99893385,\n",
       " 0.99993575,\n",
       " 0.89920354,\n",
       " 0.99843186,\n",
       " 0.9905997,\n",
       " 0.69945097,\n",
       " 0.999977,\n",
       " 1.0,\n",
       " ...]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results = bnet.model.predict(x_test)\n",
    "[np.amax(res) for res in results]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "boring-nursing",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "layer_1 (Conv2D)             (None, 28, 28, 32)        320       \n",
      "_________________________________________________________________\n",
      "max_pooling2d (MaxPooling2D) (None, 14, 14, 32)        0         \n",
      "_________________________________________________________________\n",
      "layer_2 (Conv2D)             (None, 14, 14, 64)        18496     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 7, 7, 64)          0         \n",
      "_________________________________________________________________\n",
      "layer_3 (Conv2D)             (None, 7, 7, 128)         73856     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 3, 3, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten (Flatten)            (None, 1152)              0         \n",
      "_________________________________________________________________\n",
      "dropout (Dropout)            (None, 1152)              0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 512)               590336    \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 10)                5130      \n",
      "=================================================================\n",
      "Total params: 688,138\n",
      "Trainable params: 688,138\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "bnet.model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "unique-wells",
   "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": "3.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}