10-DL-Handwritten-Digit-Recognition/Handwritten_Digit_Recognition.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "781d6368",
   "metadata": {},
   "source": [
    "# Handwritten Digit Recognition"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6f984689",
   "metadata": {},
   "source": [
    "### Importing dependencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "id": "65ec6693",
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "183d0a2f",
   "metadata": {},
   "source": [
    "### Data Collection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "id": "efc04295",
   "metadata": {},
   "outputs": [],
   "source": [
    "mnist_dataset = tf.keras.datasets.mnist"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "734c0416",
   "metadata": {},
   "source": [
    "### Train Test Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "id": "5f24a63d",
   "metadata": {},
   "outputs": [],
   "source": [
    "(x_train, y_train), (x_test, y_test) = mnist_dataset.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "id": "7fbfbbc9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "numpy.ndarray"
      ]
     },
     "execution_count": 132,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(x_train[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "id": "1ce29d1d",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[  0   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]\n",
      " [  0   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]\n",
      " [  0   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]\n",
      " [  0   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]\n",
      " [  0   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]\n",
      " [  0   0   0   0   0   0   0   0   0   0   0   0   3  18  18  18 126 136\n",
      "  175  26 166 255 247 127   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0   0  30  36  94 154 170 253 253 253 253 253\n",
      "  225 172 253 242 195  64   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0  49 238 253 253 253 253 253 253 253 253 251\n",
      "   93  82  82  56  39   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0  18 219 253 253 253 253 253 198 182 247 241\n",
      "    0   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0   0  80 156 107 253 253 205  11   0  43 154\n",
      "    0   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0   0   0  14   1 154 253  90   0   0   0   0\n",
      "    0   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0   0   0   0   0 139 253 190   2   0   0   0\n",
      "    0   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0   0   0   0   0  11 190 253  70   0   0   0\n",
      "    0   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0   0   0   0   0   0  35 241 225 160 108   1\n",
      "    0   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0   0   0   0   0   0   0  81 240 253 253 119\n",
      "   25   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0  45 186 253 253\n",
      "  150  27   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0  16  93 252\n",
      "  253 187   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   0 249\n",
      "  253 249  64   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0  46 130 183 253\n",
      "  253 207   2   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0   0   0   0   0   0  39 148 229 253 253 253\n",
      "  250 182   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0   0   0   0  24 114 221 253 253 253 253 201\n",
      "   78   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0   0   0  23  66 213 253 253 253 253 198  81   2\n",
      "    0   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0  18 171 219 253 253 253 253 195  80   9   0   0\n",
      "    0   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0  55 172 226 253 253 253 253 244 133  11   0   0   0   0\n",
      "    0   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0 136 253 253 253 212 135 132  16   0   0   0   0   0   0\n",
      "    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   0\n",
      "    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   0\n",
      "    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   0\n",
      "    0   0   0   0   0   0   0   0   0   0]]\n"
     ]
    }
   ],
   "source": [
    "print(x_train[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "id": "a6007d39",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# display image\n",
    "plt.imshow(x_train[0])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "id": "6f69bac4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# binary image B & W\n",
    "plt.imshow(x_train[0], cmap = plt.cm.binary)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "14acd73c",
   "metadata": {},
   "source": [
    "### Normalise"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "id": "4385fc29",
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = tf.keras.utils.normalize(x_train, axis = 1)\n",
    "x_test = tf.keras.utils.normalize(x_test, axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "id": "95b45312",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.00393124 0.02332955 0.02620568 0.02625207 0.17420356 0.17566281\n",
      "  0.28629534 0.05664824 0.51877786 0.71632322 0.77892406 0.89301644\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.05780486 0.06524513 0.16128198 0.22713296\n",
      "  0.22277047 0.32790981 0.36833534 0.3689874  0.34978968 0.32678448\n",
      "  0.368094   0.3747499  0.79066747 0.67980478 0.61494005 0.45002403\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.12250613 0.45858525 0.45852825 0.43408872 0.37314701\n",
      "  0.33153488 0.32790981 0.36833534 0.3689874  0.34978968 0.32420121\n",
      "  0.15214552 0.17865984 0.25626376 0.1573102  0.12298801 0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.04500225 0.4219755  0.45852825 0.43408872 0.37314701\n",
      "  0.33153488 0.32790981 0.28826244 0.26543758 0.34149427 0.31128482\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.1541463  0.28272888 0.18358693 0.37314701\n",
      "  0.33153488 0.26569767 0.01601458 0.         0.05945042 0.19891229\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.0253731  0.00171577 0.22713296\n",
      "  0.33153488 0.11664776 0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.20500962\n",
      "  0.33153488 0.24625638 0.00291174 0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.01622378\n",
      "  0.24897876 0.32790981 0.10191096 0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.04586451 0.31235677 0.32757096 0.23335172 0.14931733 0.00129164\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.10498298 0.34940902 0.3689874  0.34978968 0.15370495\n",
      "  0.04089933 0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.06551419 0.27127137 0.34978968 0.32678448\n",
      "  0.245396   0.05882702 0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.02333517 0.12857881 0.32549285\n",
      "  0.41390126 0.40743158 0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.32161793\n",
      "  0.41390126 0.54251585 0.20001074 0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.06697006 0.18959827 0.25300993 0.32678448\n",
      "  0.41390126 0.45100715 0.00625034 0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.05110617 0.19182076 0.33339444 0.3689874  0.34978968 0.32678448\n",
      "  0.40899334 0.39653769 0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.04117838 0.16813739\n",
      "  0.28960162 0.32790981 0.36833534 0.3689874  0.34978968 0.25961929\n",
      "  0.12760592 0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.04431706 0.11961607 0.36545809 0.37314701\n",
      "  0.33153488 0.32790981 0.36833534 0.28877275 0.111988   0.00258328\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.05298497 0.42752138 0.4219755  0.45852825 0.43408872 0.37314701\n",
      "  0.33153488 0.25273681 0.11646967 0.01312603 0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.37491383 0.56222061\n",
      "  0.66525569 0.63253163 0.48748768 0.45852825 0.43408872 0.359873\n",
      "  0.17428513 0.01425695 0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.92705966 0.82698729\n",
      "  0.74473314 0.63253163 0.4084877  0.24466922 0.22648107 0.02359823\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]]\n"
     ]
    }
   ],
   "source": [
    "print(x_train[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "511e799c",
   "metadata": {},
   "source": [
    "### Build NN model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "id": "52e2670d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Sequential model\n",
    "model = tf.keras.models.Sequential()\n",
    "\n",
    "# adding layers to model\n",
    "model.add(tf.keras.layers.Flatten())\n",
    "model.add(tf.keras.layers.Dense(128, activation = tf.nn.relu))\n",
    "model.add(tf.keras.layers.Dense(128, activation = tf.nn.relu))\n",
    "model.add(tf.keras.layers.Dense(10, activation = tf.nn.softmax))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "id": "3739fba2",
   "metadata": {},
   "outputs": [],
   "source": [
    "# compile model\n",
    "model.compile(optimizer = \"adam\", \n",
    "              loss = \"sparse_categorical_crossentropy\", \n",
    "              metrics = [\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "id": "341d46fc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2782 - accuracy: 0.9191 - val_loss: 0.1385 - val_accuracy: 0.9577\n",
      "Epoch 2/5\n",
      "1688/1688 [==============================] - 2s 1ms/step - loss: 0.1158 - accuracy: 0.9642 - val_loss: 0.0916 - val_accuracy: 0.9720\n",
      "Epoch 3/5\n",
      "1688/1688 [==============================] - 2s 1ms/step - loss: 0.0780 - accuracy: 0.9755 - val_loss: 0.0851 - val_accuracy: 0.9747\n",
      "Epoch 4/5\n",
      "1688/1688 [==============================] - 2s 1ms/step - loss: 0.0567 - accuracy: 0.9820 - val_loss: 0.0756 - val_accuracy: 0.9783\n",
      "Epoch 5/5\n",
      "1688/1688 [==============================] - 2s 1ms/step - loss: 0.0439 - accuracy: 0.9855 - val_loss: 0.0791 - val_accuracy: 0.9782\n"
     ]
    }
   ],
   "source": [
    "# model training\n",
    "history = model.fit(x_train, y_train, validation_split= 0.1, epochs = 5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "68817b18",
   "metadata": {},
   "source": [
    "### visualise training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "id": "7eb27ac7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x16989639280>"
      ]
     },
     "execution_count": 141,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history[\"accuracy\"])\n",
    "plt.plot(history.history[\"val_accuracy\"])\n",
    "\n",
    "plt.title(\"Model Accuray\")\n",
    "plt.ylabel(\"accuracy\")\n",
    "plt.xlabel(\"epoch\")\n",
    "\n",
    "plt.legend([\"training data\", \"validation data\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "id": "3892143c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1698aa5aac0>"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history[\"loss\"])\n",
    "plt.plot(history.history[\"val_loss\"])\n",
    "\n",
    "plt.title(\"Model Loss\")\n",
    "plt.ylabel(\"loss\")\n",
    "plt.xlabel(\"epoch\")\n",
    "\n",
    "plt.legend([\"training data\", \"validation data\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb95d844",
   "metadata": {},
   "source": [
    "### Model evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "id": "719bbec7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 0s 644us/step - loss: 0.0838 - accuracy: 0.9751\n"
     ]
    }
   ],
   "source": [
    "val_loss, val_acc = model.evaluate(x_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "id": "5561dce1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.98\n"
     ]
    }
   ],
   "source": [
    "print(\"Accuracy: {:.2f}\".format(val_acc))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c212eeea",
   "metadata": {},
   "source": [
    "### Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "id": "866a0d1a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 0s 612us/step\n"
     ]
    }
   ],
   "source": [
    "predictions = model.predict([x_test])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "id": "2dcf47df",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[2.75012901e-10 7.92941379e-10 1.49124912e-06 ... 9.99993205e-01\n",
      "  1.61914682e-09 1.08837256e-07]\n",
      " [1.07348497e-09 5.50810182e-06 9.99993920e-01 ... 1.73364559e-10\n",
      "  5.31562838e-09 3.82090227e-14]\n",
      " [2.84018000e-08 9.99933839e-01 8.10207075e-06 ... 4.83495351e-05\n",
      "  4.96033499e-06 2.66409010e-08]\n",
      " ...\n",
      " [4.35211893e-12 9.80476367e-10 3.42069773e-09 ... 2.29433817e-06\n",
      "  5.88775535e-08 1.07767182e-05]\n",
      " [2.49135127e-07 2.32065709e-08 6.81526036e-09 ... 8.21996593e-07\n",
      "  4.52756831e-05 8.16125034e-09]\n",
      " [1.32122148e-06 1.49930812e-09 1.77417917e-07 ... 9.20122919e-12\n",
      "  3.79565286e-08 4.59893235e-10]]\n"
     ]
    }
   ],
   "source": [
    "print(predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "id": "c167e9e7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4\n"
     ]
    }
   ],
   "source": [
    "print(np.argmax(predictions[4]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "id": "19366707",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# check the actual image\n",
    "plt.imshow(x_test[4])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "29a99c37",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}