NB4 NN SVM CNN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "executionInfo": {
     "elapsed": 3745,
     "status": "ok",
     "timestamp": 1700969965090,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "vFUxJY074p1K"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn import svm\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import Sequential\n",
    "from tensorflow.keras import layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 2113,
     "status": "ok",
     "timestamp": 1700969967199,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "SUOx5vGM7Bal",
    "outputId": "4593a9fa-67f3-4973-f9e6-11d78784efb5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
     ]
    }
   ],
   "source": [
    "from google.colab import drive\n",
    "drive.mount('/content/drive')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "cljw7daOVrrM"
   },
   "source": [
    "### Multiclass SVM for Wine Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "executionInfo": {
     "elapsed": 30,
     "status": "ok",
     "timestamp": 1700969967200,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "j8VE9bEj5cTO"
   },
   "outputs": [],
   "source": [
    "wine_data = np.load(\"/content/drive/My Drive/Colab Notebooks/datasets/wineData.npy\")\n",
    "wine_label = np.load(\"/content/drive/My Drive/Colab Notebooks/datasets/wineLabels.npy\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "executionInfo": {
     "elapsed": 29,
     "status": "ok",
     "timestamp": 1700969967200,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "cwCeGbqY6hp4"
   },
   "outputs": [],
   "source": [
    "x_train, x_test, y_train, y_test = train_test_split(wine_data, wine_label,\n",
    "                                                    test_size=0.2,\n",
    "                                                    random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "executionInfo": {
     "elapsed": 28,
     "status": "ok",
     "timestamp": 1700969967200,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "bUQPbZjs7Je2"
   },
   "outputs": [],
   "source": [
    "standardize = StandardScaler()\n",
    "standardized_model = standardize.fit(x_train)\n",
    "x_train_std = standardized_model.transform(x_train)\n",
    "x_test_std = standardized_model.transform(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "executionInfo": {
     "elapsed": 29,
     "status": "ok",
     "timestamp": 1700969967201,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "WVABdztd8m00"
   },
   "outputs": [],
   "source": [
    "svm_obj = svm.SVC()\n",
    "svm_model = svm_obj.fit(x_train_std, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 28,
     "status": "ok",
     "timestamp": 1700969967201,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "ZoA3aS3H-qO0",
    "outputId": "746ab9c5-2ede-42de-f5ee-7b98eb6305e7"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 100.0000%\n"
     ]
    }
   ],
   "source": [
    "pred_labels = svm_model.predict(x_test_std)\n",
    "accuracy = svm_model.score(x_test_std, y_test)\n",
    "print(f'Accuracy: {accuracy*100:0.4f}%')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 511
    },
    "executionInfo": {
     "elapsed": 1251,
     "status": "ok",
     "timestamp": 1700969968438,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "0WGCfqc__6T0",
    "outputId": "20b937b3-c704-4757-d0b3-2d2ac3beedc8"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.02, 0.5, 'Feature dimension 1')"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1300x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1, 2, figsize=(13, 5), sharex=True, sharey=True)\n",
    "ax[0].scatter(x_test_std[:, 0], x_test_std[:, 1], c=y_test)\n",
    "ax[0].set_title(\"Ground Truth\")\n",
    "ax[1].scatter(x_test_std[:, 0], x_test_std[:, 1], c=pred_labels)\n",
    "ax[1].set_title(\"Predicted Truth\")\n",
    "fig.supxlabel(\"Feature dimension 0\")\n",
    "fig.supylabel(\"Feature dimension 1\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2SqtOuEdVhTA"
   },
   "source": [
    "### Neural Network for Wine Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "executionInfo": {
     "elapsed": 1185,
     "status": "ok",
     "timestamp": 1700969969620,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "YH0Iu2wnPB_t"
   },
   "outputs": [],
   "source": [
    "model = tf.keras.Sequential(layers=[\n",
    "    tf.keras.layers.Dense(4, activation=tf.nn.relu),\n",
    "    tf.keras.layers.Dense(10, activation=tf.nn.relu),\n",
    "    tf.keras.layers.Dense(len(np.unique(y_train)), activation=tf.nn.softmax)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "executionInfo": {
     "elapsed": 11,
     "status": "ok",
     "timestamp": 1700969969620,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "ybQD2IHHq5ri"
   },
   "outputs": [],
   "source": [
    "model.compile(optimizer=tf.keras.optimizers.SGD(),\n",
    "              loss=tf.keras.losses.SparseCategoricalCrossentropy(),\n",
    "              metrics=tf.keras.metrics.SparseCategoricalAccuracy()\n",
    "              )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 26280,
     "status": "ok",
     "timestamp": 1700969995890,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "_mfFdBK4Jp8X",
    "outputId": "5e0800cc-fc12-4882-b29a-f26fbc50323f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/150\n",
      "4/4 [==============================] - 5s 295ms/step - loss: 1.1935 - sparse_categorical_accuracy: 0.3451 - val_loss: 1.2626 - val_sparse_categorical_accuracy: 0.2069\n",
      "Epoch 2/150\n",
      "4/4 [==============================] - 0s 69ms/step - loss: 1.1703 - sparse_categorical_accuracy: 0.3274 - val_loss: 1.2464 - val_sparse_categorical_accuracy: 0.2069\n",
      "Epoch 3/150\n",
      "4/4 [==============================] - 0s 51ms/step - loss: 1.1514 - sparse_categorical_accuracy: 0.3363 - val_loss: 1.2316 - val_sparse_categorical_accuracy: 0.2069\n",
      "Epoch 4/150\n",
      "4/4 [==============================] - 0s 54ms/step - loss: 1.1344 - sparse_categorical_accuracy: 0.3097 - val_loss: 1.2195 - val_sparse_categorical_accuracy: 0.1724\n",
      "Epoch 5/150\n",
      "4/4 [==============================] - 0s 42ms/step - loss: 1.1194 - sparse_categorical_accuracy: 0.2920 - val_loss: 1.2067 - val_sparse_categorical_accuracy: 0.1379\n",
      "Epoch 6/150\n",
      "4/4 [==============================] - 0s 42ms/step - loss: 1.1053 - sparse_categorical_accuracy: 0.2655 - val_loss: 1.1969 - val_sparse_categorical_accuracy: 0.1379\n",
      "Epoch 7/150\n",
      "4/4 [==============================] - 0s 38ms/step - loss: 1.0923 - sparse_categorical_accuracy: 0.2655 - val_loss: 1.1870 - val_sparse_categorical_accuracy: 0.1724\n",
      "Epoch 8/150\n",
      "4/4 [==============================] - 0s 83ms/step - loss: 1.0804 - sparse_categorical_accuracy: 0.2655 - val_loss: 1.1766 - val_sparse_categorical_accuracy: 0.2069\n",
      "Epoch 9/150\n",
      "4/4 [==============================] - 0s 33ms/step - loss: 1.0680 - sparse_categorical_accuracy: 0.2743 - val_loss: 1.1676 - val_sparse_categorical_accuracy: 0.2414\n",
      "Epoch 10/150\n",
      "4/4 [==============================] - 0s 34ms/step - loss: 1.0568 - sparse_categorical_accuracy: 0.3274 - val_loss: 1.1591 - val_sparse_categorical_accuracy: 0.2069\n",
      "Epoch 11/150\n",
      "4/4 [==============================] - 0s 46ms/step - loss: 1.0457 - sparse_categorical_accuracy: 0.4071 - val_loss: 1.1501 - val_sparse_categorical_accuracy: 0.2069\n",
      "Epoch 12/150\n",
      "4/4 [==============================] - 0s 59ms/step - loss: 1.0351 - sparse_categorical_accuracy: 0.4425 - val_loss: 1.1426 - val_sparse_categorical_accuracy: 0.2759\n",
      "Epoch 13/150\n",
      "4/4 [==============================] - 0s 60ms/step - loss: 1.0247 - sparse_categorical_accuracy: 0.4779 - val_loss: 1.1356 - val_sparse_categorical_accuracy: 0.3103\n",
      "Epoch 14/150\n",
      "4/4 [==============================] - 0s 46ms/step - loss: 1.0144 - sparse_categorical_accuracy: 0.4956 - val_loss: 1.1291 - val_sparse_categorical_accuracy: 0.3448\n",
      "Epoch 15/150\n",
      "4/4 [==============================] - 0s 65ms/step - loss: 1.0041 - sparse_categorical_accuracy: 0.5044 - val_loss: 1.1213 - val_sparse_categorical_accuracy: 0.3793\n",
      "Epoch 16/150\n",
      "4/4 [==============================] - 0s 60ms/step - loss: 0.9940 - sparse_categorical_accuracy: 0.4956 - val_loss: 1.1152 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 17/150\n",
      "4/4 [==============================] - 0s 56ms/step - loss: 0.9842 - sparse_categorical_accuracy: 0.5310 - val_loss: 1.1084 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 18/150\n",
      "4/4 [==============================] - 0s 63ms/step - loss: 0.9738 - sparse_categorical_accuracy: 0.5398 - val_loss: 1.1023 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 19/150\n",
      "4/4 [==============================] - 0s 58ms/step - loss: 0.9642 - sparse_categorical_accuracy: 0.5398 - val_loss: 1.0952 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 20/150\n",
      "4/4 [==============================] - 0s 66ms/step - loss: 0.9540 - sparse_categorical_accuracy: 0.5487 - val_loss: 1.0886 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 21/150\n",
      "4/4 [==============================] - 0s 72ms/step - loss: 0.9431 - sparse_categorical_accuracy: 0.5752 - val_loss: 1.0816 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 22/150\n",
      "4/4 [==============================] - 0s 68ms/step - loss: 0.9325 - sparse_categorical_accuracy: 0.5841 - val_loss: 1.0744 - val_sparse_categorical_accuracy: 0.3793\n",
      "Epoch 23/150\n",
      "4/4 [==============================] - 0s 42ms/step - loss: 0.9218 - sparse_categorical_accuracy: 0.5752 - val_loss: 1.0684 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 24/150\n",
      "4/4 [==============================] - 0s 69ms/step - loss: 0.9110 - sparse_categorical_accuracy: 0.6106 - val_loss: 1.0617 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 25/150\n",
      "4/4 [==============================] - 0s 63ms/step - loss: 0.9014 - sparse_categorical_accuracy: 0.6195 - val_loss: 1.0542 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 26/150\n",
      "4/4 [==============================] - 0s 43ms/step - loss: 0.8892 - sparse_categorical_accuracy: 0.6460 - val_loss: 1.0461 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 27/150\n",
      "4/4 [==============================] - 0s 61ms/step - loss: 0.8779 - sparse_categorical_accuracy: 0.6372 - val_loss: 1.0390 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 28/150\n",
      "4/4 [==============================] - 0s 50ms/step - loss: 0.8664 - sparse_categorical_accuracy: 0.6372 - val_loss: 1.0310 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 29/150\n",
      "4/4 [==============================] - 0s 68ms/step - loss: 0.8544 - sparse_categorical_accuracy: 0.6460 - val_loss: 1.0228 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 30/150\n",
      "4/4 [==============================] - 0s 69ms/step - loss: 0.8430 - sparse_categorical_accuracy: 0.6549 - val_loss: 1.0147 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 31/150\n",
      "4/4 [==============================] - 0s 41ms/step - loss: 0.8315 - sparse_categorical_accuracy: 0.6726 - val_loss: 1.0058 - val_sparse_categorical_accuracy: 0.4138\n",
      "Epoch 32/150\n",
      "4/4 [==============================] - 0s 63ms/step - loss: 0.8201 - sparse_categorical_accuracy: 0.6726 - val_loss: 0.9979 - val_sparse_categorical_accuracy: 0.4483\n",
      "Epoch 33/150\n",
      "4/4 [==============================] - 0s 47ms/step - loss: 0.8089 - sparse_categorical_accuracy: 0.6726 - val_loss: 0.9900 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 34/150\n",
      "4/4 [==============================] - 0s 35ms/step - loss: 0.7966 - sparse_categorical_accuracy: 0.6814 - val_loss: 0.9812 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 35/150\n",
      "4/4 [==============================] - 0s 33ms/step - loss: 0.7854 - sparse_categorical_accuracy: 0.6903 - val_loss: 0.9729 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 36/150\n",
      "4/4 [==============================] - 0s 28ms/step - loss: 0.7736 - sparse_categorical_accuracy: 0.6991 - val_loss: 0.9644 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 37/150\n",
      "4/4 [==============================] - 0s 52ms/step - loss: 0.7605 - sparse_categorical_accuracy: 0.6991 - val_loss: 0.9555 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 38/150\n",
      "4/4 [==============================] - 0s 36ms/step - loss: 0.7484 - sparse_categorical_accuracy: 0.7080 - val_loss: 0.9471 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 39/150\n",
      "4/4 [==============================] - 0s 35ms/step - loss: 0.7370 - sparse_categorical_accuracy: 0.7168 - val_loss: 0.9389 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 40/150\n",
      "4/4 [==============================] - 0s 43ms/step - loss: 0.7253 - sparse_categorical_accuracy: 0.7168 - val_loss: 0.9309 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 41/150\n",
      "4/4 [==============================] - 0s 28ms/step - loss: 0.7134 - sparse_categorical_accuracy: 0.7168 - val_loss: 0.9222 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 42/150\n",
      "4/4 [==============================] - 0s 35ms/step - loss: 0.7022 - sparse_categorical_accuracy: 0.7168 - val_loss: 0.9141 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 43/150\n",
      "4/4 [==============================] - 0s 29ms/step - loss: 0.6912 - sparse_categorical_accuracy: 0.7257 - val_loss: 0.9056 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 44/150\n",
      "4/4 [==============================] - 0s 25ms/step - loss: 0.6802 - sparse_categorical_accuracy: 0.7345 - val_loss: 0.8970 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 45/150\n",
      "4/4 [==============================] - 0s 38ms/step - loss: 0.6692 - sparse_categorical_accuracy: 0.7345 - val_loss: 0.8877 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 46/150\n",
      "4/4 [==============================] - 0s 44ms/step - loss: 0.6582 - sparse_categorical_accuracy: 0.7345 - val_loss: 0.8773 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 47/150\n",
      "4/4 [==============================] - 0s 35ms/step - loss: 0.6473 - sparse_categorical_accuracy: 0.7257 - val_loss: 0.8674 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 48/150\n",
      "4/4 [==============================] - 0s 20ms/step - loss: 0.6363 - sparse_categorical_accuracy: 0.7434 - val_loss: 0.8582 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 49/150\n",
      "4/4 [==============================] - 0s 25ms/step - loss: 0.6260 - sparse_categorical_accuracy: 0.7434 - val_loss: 0.8489 - val_sparse_categorical_accuracy: 0.4828\n",
      "Epoch 50/150\n",
      "4/4 [==============================] - 0s 48ms/step - loss: 0.6162 - sparse_categorical_accuracy: 0.7345 - val_loss: 0.8398 - val_sparse_categorical_accuracy: 0.5172\n",
      "Epoch 51/150\n",
      "4/4 [==============================] - 0s 46ms/step - loss: 0.6062 - sparse_categorical_accuracy: 0.7434 - val_loss: 0.8314 - val_sparse_categorical_accuracy: 0.5172\n",
      "Epoch 52/150\n",
      "4/4 [==============================] - 0s 26ms/step - loss: 0.5963 - sparse_categorical_accuracy: 0.7522 - val_loss: 0.8222 - val_sparse_categorical_accuracy: 0.5517\n",
      "Epoch 53/150\n",
      "4/4 [==============================] - 0s 28ms/step - loss: 0.5861 - sparse_categorical_accuracy: 0.7611 - val_loss: 0.8139 - val_sparse_categorical_accuracy: 0.5517\n",
      "Epoch 54/150\n",
      "4/4 [==============================] - 0s 39ms/step - loss: 0.5765 - sparse_categorical_accuracy: 0.7522 - val_loss: 0.8051 - val_sparse_categorical_accuracy: 0.5517\n",
      "Epoch 55/150\n",
      "4/4 [==============================] - 0s 37ms/step - loss: 0.5676 - sparse_categorical_accuracy: 0.7611 - val_loss: 0.7959 - val_sparse_categorical_accuracy: 0.5862\n",
      "Epoch 56/150\n",
      "4/4 [==============================] - 0s 28ms/step - loss: 0.5569 - sparse_categorical_accuracy: 0.7611 - val_loss: 0.7871 - val_sparse_categorical_accuracy: 0.5862\n",
      "Epoch 57/150\n",
      "4/4 [==============================] - 0s 26ms/step - loss: 0.5484 - sparse_categorical_accuracy: 0.7522 - val_loss: 0.7776 - val_sparse_categorical_accuracy: 0.6207\n",
      "Epoch 58/150\n",
      "4/4 [==============================] - 0s 40ms/step - loss: 0.5385 - sparse_categorical_accuracy: 0.7611 - val_loss: 0.7685 - val_sparse_categorical_accuracy: 0.6207\n",
      "Epoch 59/150\n",
      "4/4 [==============================] - 0s 26ms/step - loss: 0.5295 - sparse_categorical_accuracy: 0.7699 - val_loss: 0.7589 - val_sparse_categorical_accuracy: 0.7241\n",
      "Epoch 60/150\n",
      "4/4 [==============================] - 0s 25ms/step - loss: 0.5205 - sparse_categorical_accuracy: 0.8053 - val_loss: 0.7497 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 61/150\n",
      "4/4 [==============================] - 0s 28ms/step - loss: 0.5118 - sparse_categorical_accuracy: 0.8319 - val_loss: 0.7403 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 62/150\n",
      "4/4 [==============================] - 0s 29ms/step - loss: 0.5027 - sparse_categorical_accuracy: 0.8407 - val_loss: 0.7322 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 63/150\n",
      "4/4 [==============================] - 0s 26ms/step - loss: 0.4944 - sparse_categorical_accuracy: 0.8584 - val_loss: 0.7236 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 64/150\n",
      "4/4 [==============================] - 0s 27ms/step - loss: 0.4867 - sparse_categorical_accuracy: 0.8673 - val_loss: 0.7156 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 65/150\n",
      "4/4 [==============================] - 0s 31ms/step - loss: 0.4785 - sparse_categorical_accuracy: 0.8761 - val_loss: 0.7082 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 66/150\n",
      "4/4 [==============================] - 0s 22ms/step - loss: 0.4707 - sparse_categorical_accuracy: 0.8850 - val_loss: 0.7002 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 67/150\n",
      "4/4 [==============================] - 0s 30ms/step - loss: 0.4632 - sparse_categorical_accuracy: 0.8850 - val_loss: 0.6926 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 68/150\n",
      "4/4 [==============================] - 0s 22ms/step - loss: 0.4560 - sparse_categorical_accuracy: 0.8938 - val_loss: 0.6847 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 69/150\n",
      "4/4 [==============================] - 0s 40ms/step - loss: 0.4489 - sparse_categorical_accuracy: 0.8938 - val_loss: 0.6771 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 70/150\n",
      "4/4 [==============================] - 0s 35ms/step - loss: 0.4418 - sparse_categorical_accuracy: 0.8938 - val_loss: 0.6698 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 71/150\n",
      "4/4 [==============================] - 0s 34ms/step - loss: 0.4350 - sparse_categorical_accuracy: 0.8938 - val_loss: 0.6623 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 72/150\n",
      "4/4 [==============================] - 0s 36ms/step - loss: 0.4282 - sparse_categorical_accuracy: 0.9027 - val_loss: 0.6553 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 73/150\n",
      "4/4 [==============================] - 0s 27ms/step - loss: 0.4221 - sparse_categorical_accuracy: 0.9027 - val_loss: 0.6466 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 74/150\n",
      "4/4 [==============================] - 0s 25ms/step - loss: 0.4152 - sparse_categorical_accuracy: 0.9027 - val_loss: 0.6394 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 75/150\n",
      "4/4 [==============================] - 0s 24ms/step - loss: 0.4093 - sparse_categorical_accuracy: 0.9027 - val_loss: 0.6318 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 76/150\n",
      "4/4 [==============================] - 0s 23ms/step - loss: 0.4030 - sparse_categorical_accuracy: 0.9027 - val_loss: 0.6239 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 77/150\n",
      "4/4 [==============================] - 0s 23ms/step - loss: 0.3968 - sparse_categorical_accuracy: 0.9027 - val_loss: 0.6152 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 78/150\n",
      "4/4 [==============================] - 0s 44ms/step - loss: 0.3909 - sparse_categorical_accuracy: 0.9027 - val_loss: 0.6071 - val_sparse_categorical_accuracy: 0.8276\n",
      "Epoch 79/150\n",
      "4/4 [==============================] - 0s 22ms/step - loss: 0.3842 - sparse_categorical_accuracy: 0.9027 - val_loss: 0.5990 - val_sparse_categorical_accuracy: 0.8276\n",
      "Epoch 80/150\n",
      "4/4 [==============================] - 0s 29ms/step - loss: 0.3782 - sparse_categorical_accuracy: 0.9027 - val_loss: 0.5910 - val_sparse_categorical_accuracy: 0.8276\n",
      "Epoch 81/150\n",
      "4/4 [==============================] - 0s 26ms/step - loss: 0.3725 - sparse_categorical_accuracy: 0.9027 - val_loss: 0.5834 - val_sparse_categorical_accuracy: 0.8276\n",
      "Epoch 82/150\n",
      "4/4 [==============================] - 0s 28ms/step - loss: 0.3671 - sparse_categorical_accuracy: 0.9027 - val_loss: 0.5751 - val_sparse_categorical_accuracy: 0.8276\n",
      "Epoch 83/150\n",
      "4/4 [==============================] - 0s 23ms/step - loss: 0.3615 - sparse_categorical_accuracy: 0.9027 - val_loss: 0.5665 - val_sparse_categorical_accuracy: 0.8276\n",
      "Epoch 84/150\n",
      "4/4 [==============================] - 0s 25ms/step - loss: 0.3557 - sparse_categorical_accuracy: 0.9027 - val_loss: 0.5588 - val_sparse_categorical_accuracy: 0.8276\n",
      "Epoch 85/150\n",
      "4/4 [==============================] - 0s 24ms/step - loss: 0.3505 - sparse_categorical_accuracy: 0.9115 - val_loss: 0.5512 - val_sparse_categorical_accuracy: 0.8276\n",
      "Epoch 86/150\n",
      "4/4 [==============================] - 0s 30ms/step - loss: 0.3452 - sparse_categorical_accuracy: 0.9115 - val_loss: 0.5429 - val_sparse_categorical_accuracy: 0.8276\n",
      "Epoch 87/150\n",
      "4/4 [==============================] - 0s 22ms/step - loss: 0.3398 - sparse_categorical_accuracy: 0.9115 - val_loss: 0.5363 - val_sparse_categorical_accuracy: 0.8276\n",
      "Epoch 88/150\n",
      "4/4 [==============================] - 0s 22ms/step - loss: 0.3347 - sparse_categorical_accuracy: 0.9115 - val_loss: 0.5292 - val_sparse_categorical_accuracy: 0.8621\n",
      "Epoch 89/150\n",
      "4/4 [==============================] - 0s 37ms/step - loss: 0.3298 - sparse_categorical_accuracy: 0.9115 - val_loss: 0.5225 - val_sparse_categorical_accuracy: 0.8621\n",
      "Epoch 90/150\n",
      "4/4 [==============================] - 0s 17ms/step - loss: 0.3249 - sparse_categorical_accuracy: 0.9292 - val_loss: 0.5163 - val_sparse_categorical_accuracy: 0.8621\n",
      "Epoch 91/150\n",
      "4/4 [==============================] - 0s 31ms/step - loss: 0.3204 - sparse_categorical_accuracy: 0.9292 - val_loss: 0.5100 - val_sparse_categorical_accuracy: 0.8621\n",
      "Epoch 92/150\n",
      "4/4 [==============================] - 0s 29ms/step - loss: 0.3157 - sparse_categorical_accuracy: 0.9292 - val_loss: 0.5025 - val_sparse_categorical_accuracy: 0.8621\n",
      "Epoch 93/150\n",
      "4/4 [==============================] - 0s 36ms/step - loss: 0.3106 - sparse_categorical_accuracy: 0.9292 - val_loss: 0.4954 - val_sparse_categorical_accuracy: 0.8621\n",
      "Epoch 94/150\n",
      "4/4 [==============================] - 0s 27ms/step - loss: 0.3063 - sparse_categorical_accuracy: 0.9292 - val_loss: 0.4885 - val_sparse_categorical_accuracy: 0.8621\n",
      "Epoch 95/150\n",
      "4/4 [==============================] - 0s 50ms/step - loss: 0.3016 - sparse_categorical_accuracy: 0.9292 - val_loss: 0.4811 - val_sparse_categorical_accuracy: 0.8966\n",
      "Epoch 96/150\n",
      "4/4 [==============================] - 0s 28ms/step - loss: 0.2971 - sparse_categorical_accuracy: 0.9292 - val_loss: 0.4733 - val_sparse_categorical_accuracy: 0.8966\n",
      "Epoch 97/150\n",
      "4/4 [==============================] - 0s 31ms/step - loss: 0.2925 - sparse_categorical_accuracy: 0.9292 - val_loss: 0.4677 - val_sparse_categorical_accuracy: 0.8966\n",
      "Epoch 98/150\n",
      "4/4 [==============================] - 0s 23ms/step - loss: 0.2880 - sparse_categorical_accuracy: 0.9292 - val_loss: 0.4613 - val_sparse_categorical_accuracy: 0.8966\n",
      "Epoch 99/150\n",
      "4/4 [==============================] - 0s 23ms/step - loss: 0.2836 - sparse_categorical_accuracy: 0.9381 - val_loss: 0.4537 - val_sparse_categorical_accuracy: 0.8966\n",
      "Epoch 100/150\n",
      "4/4 [==============================] - 0s 25ms/step - loss: 0.2790 - sparse_categorical_accuracy: 0.9381 - val_loss: 0.4477 - val_sparse_categorical_accuracy: 0.8966\n",
      "Epoch 101/150\n",
      "4/4 [==============================] - 0s 28ms/step - loss: 0.2749 - sparse_categorical_accuracy: 0.9469 - val_loss: 0.4412 - val_sparse_categorical_accuracy: 0.8966\n",
      "Epoch 102/150\n",
      "4/4 [==============================] - 0s 26ms/step - loss: 0.2706 - sparse_categorical_accuracy: 0.9469 - val_loss: 0.4360 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 103/150\n",
      "4/4 [==============================] - 0s 39ms/step - loss: 0.2667 - sparse_categorical_accuracy: 0.9469 - val_loss: 0.4300 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 104/150\n",
      "4/4 [==============================] - 0s 26ms/step - loss: 0.2625 - sparse_categorical_accuracy: 0.9469 - val_loss: 0.4240 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 105/150\n",
      "4/4 [==============================] - 0s 27ms/step - loss: 0.2587 - sparse_categorical_accuracy: 0.9469 - val_loss: 0.4188 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 106/150\n",
      "4/4 [==============================] - 0s 76ms/step - loss: 0.2548 - sparse_categorical_accuracy: 0.9469 - val_loss: 0.4137 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 107/150\n",
      "4/4 [==============================] - 0s 62ms/step - loss: 0.2509 - sparse_categorical_accuracy: 0.9469 - val_loss: 0.4077 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 108/150\n",
      "4/4 [==============================] - 0s 47ms/step - loss: 0.2473 - sparse_categorical_accuracy: 0.9469 - val_loss: 0.4015 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 109/150\n",
      "4/4 [==============================] - 0s 60ms/step - loss: 0.2437 - sparse_categorical_accuracy: 0.9558 - val_loss: 0.3964 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 110/150\n",
      "4/4 [==============================] - 0s 28ms/step - loss: 0.2401 - sparse_categorical_accuracy: 0.9558 - val_loss: 0.3908 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 111/150\n",
      "4/4 [==============================] - 0s 33ms/step - loss: 0.2366 - sparse_categorical_accuracy: 0.9558 - val_loss: 0.3859 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 112/150\n",
      "4/4 [==============================] - 0s 27ms/step - loss: 0.2331 - sparse_categorical_accuracy: 0.9558 - val_loss: 0.3816 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 113/150\n",
      "4/4 [==============================] - 0s 23ms/step - loss: 0.2298 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3769 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 114/150\n",
      "4/4 [==============================] - 0s 27ms/step - loss: 0.2266 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3723 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 115/150\n",
      "4/4 [==============================] - 0s 24ms/step - loss: 0.2235 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3675 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 116/150\n",
      "4/4 [==============================] - 0s 21ms/step - loss: 0.2206 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3631 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 117/150\n",
      "4/4 [==============================] - 0s 54ms/step - loss: 0.2176 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3593 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 118/150\n",
      "4/4 [==============================] - 0s 30ms/step - loss: 0.2147 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3548 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 119/150\n",
      "4/4 [==============================] - 0s 29ms/step - loss: 0.2118 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3502 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 120/150\n",
      "4/4 [==============================] - 0s 44ms/step - loss: 0.2088 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3463 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 121/150\n",
      "4/4 [==============================] - 0s 68ms/step - loss: 0.2061 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3425 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 122/150\n",
      "4/4 [==============================] - 0s 43ms/step - loss: 0.2036 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3380 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 123/150\n",
      "4/4 [==============================] - 0s 54ms/step - loss: 0.2011 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3337 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 124/150\n",
      "4/4 [==============================] - 0s 71ms/step - loss: 0.1985 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3290 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 125/150\n",
      "4/4 [==============================] - 0s 74ms/step - loss: 0.1959 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3248 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 126/150\n",
      "4/4 [==============================] - 0s 55ms/step - loss: 0.1934 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3215 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 127/150\n",
      "4/4 [==============================] - 0s 47ms/step - loss: 0.1912 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3193 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 128/150\n",
      "4/4 [==============================] - 0s 46ms/step - loss: 0.1890 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3154 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 129/150\n",
      "4/4 [==============================] - 0s 50ms/step - loss: 0.1866 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3120 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 130/150\n",
      "4/4 [==============================] - 0s 49ms/step - loss: 0.1845 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3090 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 131/150\n",
      "4/4 [==============================] - 0s 48ms/step - loss: 0.1823 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3054 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 132/150\n",
      "4/4 [==============================] - 0s 47ms/step - loss: 0.1802 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.3024 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 133/150\n",
      "4/4 [==============================] - 0s 58ms/step - loss: 0.1781 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2987 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 134/150\n",
      "4/4 [==============================] - 0s 25ms/step - loss: 0.1759 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2958 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 135/150\n",
      "4/4 [==============================] - 0s 47ms/step - loss: 0.1741 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2928 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 136/150\n",
      "4/4 [==============================] - 0s 31ms/step - loss: 0.1721 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2892 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 137/150\n",
      "4/4 [==============================] - 0s 37ms/step - loss: 0.1700 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2865 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 138/150\n",
      "4/4 [==============================] - 0s 49ms/step - loss: 0.1682 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2842 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 139/150\n",
      "4/4 [==============================] - 0s 73ms/step - loss: 0.1664 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2814 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 140/150\n",
      "4/4 [==============================] - 0s 65ms/step - loss: 0.1647 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2790 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 141/150\n",
      "4/4 [==============================] - 0s 63ms/step - loss: 0.1630 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2768 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 142/150\n",
      "4/4 [==============================] - 0s 68ms/step - loss: 0.1616 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2742 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 143/150\n",
      "4/4 [==============================] - 0s 36ms/step - loss: 0.1600 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2713 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 144/150\n",
      "4/4 [==============================] - 0s 53ms/step - loss: 0.1583 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2685 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 145/150\n",
      "4/4 [==============================] - 0s 38ms/step - loss: 0.1566 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2662 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 146/150\n",
      "4/4 [==============================] - 0s 36ms/step - loss: 0.1550 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2637 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 147/150\n",
      "4/4 [==============================] - 0s 38ms/step - loss: 0.1535 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2619 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 148/150\n",
      "4/4 [==============================] - 0s 45ms/step - loss: 0.1520 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2589 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 149/150\n",
      "4/4 [==============================] - 0s 38ms/step - loss: 0.1504 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2564 - val_sparse_categorical_accuracy: 0.9310\n",
      "Epoch 150/150\n",
      "4/4 [==============================] - 0s 45ms/step - loss: 0.1488 - sparse_categorical_accuracy: 0.9646 - val_loss: 0.2536 - val_sparse_categorical_accuracy: 0.9310\n"
     ]
    }
   ],
   "source": [
    "# Stadardized data from Task 4.1\n",
    "history = model.fit(x_train_std, y_train, epochs=150, validation_split=0.2, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 23,
     "status": "ok",
     "timestamp": 1700969995891,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "eikl9_edJqDF",
    "outputId": "640f3d83-ed72-4109-9e04-4c13f2545198"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense (Dense)               (None, 4)                 56        \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 10)                50        \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 3)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 139 (556.00 Byte)\n",
      "Trainable params: 139 (556.00 Byte)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 611,
     "status": "ok",
     "timestamp": 1700969996496,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "yKwHH2jGux3z",
    "outputId": "ff4f2917-788c-4149-d066-e7b82c39c67b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2/2 [==============================] - 0s 6ms/step\n"
     ]
    }
   ],
   "source": [
    "pred_labels_nn = model.predict(x_test_std)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 511
    },
    "executionInfo": {
     "elapsed": 1952,
     "status": "ok",
     "timestamp": 1700969998440,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "Pxk1MfKbt91G",
    "outputId": "640ceaa6-01dc-4e0a-b665-c4408ee04349"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.02, 0.5, 'Feature dimension 1')"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1300x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig1, ax1 = plt.subplots(1, 2, figsize=(13, 5), sharex=True, sharey=True)\n",
    "ax1[0].scatter(x_test_std[:, 0], x_test_std[:, 1], c=y_test)\n",
    "ax1[0].set_title(\"Ground Truth\")\n",
    "ax1[1].scatter(x_test_std[:, 0], x_test_std[:, 1], c=pred_labels_nn)\n",
    "ax1[1].set_title(\"Predicted Truth\")\n",
    "fig1.supxlabel(\"Feature dimension 0\")\n",
    "fig1.supylabel(\"Feature dimension 1\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 504
    },
    "executionInfo": {
     "elapsed": 1119,
     "status": "ok",
     "timestamp": 1700969999550,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "CvocxNtF3Ee8",
    "outputId": "9da50dd9-8547-4787-ede2-35819193c559"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Loss Curves')"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1300x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig2, ax2 = plt.subplots(1, 2, figsize=(13, 5))\n",
    "\n",
    "ax2[0].plot(history.history['sparse_categorical_accuracy'])\n",
    "ax2[0].plot(history.history['val_sparse_categorical_accuracy'])\n",
    "ax2[0].set_xlabel('Epoch')\n",
    "ax2[0].set_ylabel('Accuracy')\n",
    "ax2[0].legend(['Training Accuracy', 'Validation Accuracy'], loc='lower right')\n",
    "ax2[0].set_title('Accuracy Curves')\n",
    "\n",
    "ax2[1].plot(history.history['loss'])\n",
    "ax2[1].plot(history.history['val_loss'])\n",
    "\n",
    "ax2[1].set_xlabel('Epoch')\n",
    "ax2[1].set_ylabel('Loss')\n",
    "ax2[1].legend(['Training Loss', 'Validation Loss'])\n",
    "ax2[1].set_title('Loss Curves')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 15,
     "status": "ok",
     "timestamp": 1700969999550,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "G9LGZdBCJqKX",
    "outputId": "71a12f16-0241-46f9-c0a6-753651eaeee3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2/2 [==============================] - 0s 8ms/step - loss: 0.3078 - sparse_categorical_accuracy: 0.9167\n",
      "Test set loss: 0.31, test set accuracy: 91.67%\n"
     ]
    }
   ],
   "source": [
    "results = model.evaluate(x_test_std, y_test)\n",
    "print(f'Test set loss: {results[0]:0.2f}, test set accuracy: {results[1]*100:0.2f}%')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "VpuxS88nvqYW"
   },
   "source": [
    "### Neural Network for Digit Recognition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "executionInfo": {
     "elapsed": 621,
     "status": "ok",
     "timestamp": 1700970000163,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "7O55Uq6SJqM7"
   },
   "outputs": [],
   "source": [
    "(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 22,
     "status": "ok",
     "timestamp": 1700970000163,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "gLP0NzAVJqPK",
    "outputId": "3f22401a-6e9d-44d2-99f1-94715236ce72"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MNIST sample shape: (28, 28)\n",
      "Number of training samples: 60000\n",
      "Number of testing samples: 10000\n",
      "Number of classes: 10\n"
     ]
    }
   ],
   "source": [
    "print(f'MNIST sample shape: {x_train.shape[1:]}')\n",
    "print(f'Number of training samples: {x_train.shape[0]}')\n",
    "print(f'Number of testing samples: {x_test.shape[0]}')\n",
    "print(f'Number of classes: {len(np.unique(y_train))}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 447
    },
    "executionInfo": {
     "elapsed": 20,
     "status": "ok",
     "timestamp": 1700970000164,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "bL0W967aKJbC",
    "outputId": "f9fc14b9-7070-4e18-8d7d-415385f31281"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x789cd0b46110>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAa4AAAGdCAYAAABKG5eZAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAbfklEQVR4nO3dfWyV9f3/8dfhpgfU9nS19G4UbPGGRW6WoXSNijiali5xovwB3iRgjEQsOuycpk7Bu6UbJs65VEzmQjWKqItA5A8MVlvmVjCghBC1oV03IKUFWXoOFCmEfr5/+PP8PNIWzuU5vfrueT6SK+Gcc316vb1yJU9Pe3o14JxzAgDAiFF+DwAAQDwIFwDAFMIFADCFcAEATCFcAABTCBcAwBTCBQAwhXABAEwZ4/cA39fX16eOjg6lp6crEAj4PQ4AYIg453T8+HEVFBRo1KiB31cNu3B1dHSosLDQ7zEAAD45ePCgJk6cOODrw+5bhenp6X6PAADw0fk6kLRw1dXV6bLLLtO4ceNUUlKiTz755ILW8e1BAEht5+tAUsL11ltvqbq6WqtXr9ann36qmTNnqqKiQkeOHEnG4QAAqcQlwezZs11VVVX08dmzZ11BQYGrra0979pwOOwksbGxsbGl6BYOhwftRMLfcZ0+fVq7d+9WWVlZ9LlRo0aprKxMzc3NiT4cACDFJPxThV999ZXOnj2r3NzcmOdzc3P15ZdfnrN/b2+vent7o48jkUiiRwIAjCC+f6qwtrZWoVAouvFReADAYBIeruzsbI0ePVpdXV0xz3d1dSkvL++c/WtqahQOh6PbwYMHEz0SAGAESXi40tLSNGvWLDU0NESf6+vrU0NDg0pLS8/ZPxgMKiMjI2YDAGAgSblzRnV1tZYsWaJrrrlGs2fP1gsvvKCenh7dfffdyTgcACCFJCVcixYt0tGjR7Vq1Sp1dnbqpz/9qbZu3XrOBzYAAIhXwDnn/B7iuyKRiEKhkN9jAAB8Eg6HB/2xke+fKgQAIB6ECwBgCuECAJhCuAAAphAuAIAphAsAYArhAgCYQrgAAKYQLgCAKYQLAGAK4QIAmEK4AACmEC4AgCmECwBgCuECAJhCuAAAphAuAIAphAsAYArhAgCYQrgAAKYQLgCAKYQLAGAK4QIAmEK4AACmEC4AgCmECwBgCuECAJhCuAAAphAuAIAphAsAYArhAgCYQrgAAKYQLgCAKYQLAGAK4QIAmEK4AACmEC4AgCmECwBgCuECAJhCuAAAphAuAIAphAsAYArhAgCYQrgAAKYQLgCAKYQLAGAK4QIAmEK4AACmEC4AgCmECwBgCuECAJhCuAAAphAuAIAphAsAYArhAgCYQrgAAKYQLgCAKYQLAGBKwsP15JNPKhAIxGxTp05N9GEAAClqTDK+6NVXX60PPvjg/x9kTFIOAwBIQUkpypgxY5SXl5eMLw0ASHFJ+RnX/v37VVBQoOLiYt155506cODAgPv29vYqEonEbAAADCTh4SopKVF9fb22bt2qtWvXqr29XTfccIOOHz/e7/61tbUKhULRrbCwMNEjAQBGkIBzziXzAN3d3Zo8ebKef/553XPPPee83tvbq97e3ujjSCRCvAAghYXDYWVkZAz4etI/NZGZmakrr7xSra2t/b4eDAYVDAaTPQYAYIRI+u9xnThxQm1tbcrPz0/2oQAAKSDh4Xr44YfV1NSk//znP/rXv/6lW2+9VaNHj9btt9+e6EMBAFJQwr9VeOjQId1+++06duyYJkyYoOuvv147duzQhAkTEn0oAEAKSvqHM+IViUQUCoX8HgOIUVJSEveau+66y9OxbrzxRk/rrr76ak/rvHj44Yc9revo6PC07vrrr497zeuvv+7pWDt37vS0Dolzvg9ncK9CAIAphAsAYArhAgCYQrgAAKYQLgCAKYQLAGAK4QIAmEK4AACmEC4AgCmECwBgCuECAJhCuAAAphAuAIApSf8LyMBwsmjRIk/r/vznP8e9Jjs729OxAoGAp3WNjY1xr/H654aee+45T+u88nJOvP63LV682NM6DB3ecQEATCFcAABTCBcAwBTCBQAwhXABAEwhXAAAUwgXAMAUwgUAMIVwAQBMIVwAAFMIFwDAFMIFADCFcAEATOHu8PDVmDHeLsFrrrnG07q//vWvntZddNFFca/Zvn27p2M988wzntZ9/PHHca8JBoOejvX22297WldeXu5pnRe7du0asmNhaPGOCwBgCuECAJhCuAAAphAuAIAphAsAYArhAgCYQrgAAKYQLgCAKYQLAGAK4QIAmEK4AACmEC4AgCncZBe+uuuuuzyte+WVVxI8yeC2bdsW95pFixZ5OlYkEvG0zguvMw7lzXIl6dChQ3GvefXVV5MwCYYD3nEBAEwhXAAAUwgXAMAUwgUAMIVwAQBMIVwAAFMIFwDAFMIFADCFcAEATCFcAABTCBcAwBTCBQAwhXABAEzh7vBImGeeeSbuNY899pinYznnPK176aWXPK17/PHH414zlHd59+p3v/ud3yNckAcffDDuNUePHk3CJBgOeMcFADCFcAEATIk7XNu3b9fNN9+sgoICBQIBbdq0KeZ155xWrVql/Px8jR8/XmVlZdq/f3+i5gUApLi4w9XT06OZM2eqrq6u39fXrFmjF198US+//LJ27typiy++WBUVFTp16tQPHhYAgLg/nFFZWanKysp+X3PO6YUXXtDjjz+uW265RZL02muvKTc3V5s2bdLixYt/2LQAgJSX0J9xtbe3q7OzU2VlZdHnQqGQSkpK1Nzc3O+a3t5eRSKRmA0AgIEkNFydnZ2SpNzc3Jjnc3Nzo699X21trUKhUHQrLCxM5EgAgBHG908V1tTUKBwOR7eDBw/6PRIAYBhLaLjy8vIkSV1dXTHPd3V1RV/7vmAwqIyMjJgNAICBJDRcRUVFysvLU0NDQ/S5SCSinTt3qrS0NJGHAgCkqLg/VXjixAm1trZGH7e3t2vPnj3KysrSpEmTtHLlSj377LO64oorVFRUpCeeeEIFBQVasGBBIucGAKSouMO1a9cu3XTTTdHH1dXVkqQlS5aovr5ejzzyiHp6erRs2TJ1d3fr+uuv19atWzVu3LjETQ0ASFlxh2vu3LmD3uA0EAjo6aef1tNPP/2DBgMAoD/cHR7nWLVqlad1Xu70fvr0aU/Hev/99z2te/TRRz2t+/rrrz2t88LrdyfKy8vjXjNp0iRPxwoEAp7WPfvss57Wbd682dM6jEy+fxweAIB4EC4AgCmECwBgCuECAJhCuAAAphAuAIAphAsAYArhAgCYQrgAAKYQLgCAKYQLAGAK4QIAmMJNdkewzMxMT+vuv/9+T+sG+6sBA/F6s1wLf9/t8ssv97TujTfe8LRu1qxZntZ58fe//93TujVr1iR4EqQi3nEBAEwhXAAAUwgXAMAUwgUAMIVwAQBMIVwAAFMIFwDAFMIFADCFcAEATCFcAABTCBcAwBTCBQAwhXABAEzh7vAjWFpamqd12dnZCZ5kYA8++KCndTk5OZ7W3X333Z7W/epXv4p7zbRp0zwd65JLLvG0zsvd+b2skaTXX3/d07qenh5P64Dv4h0XAMAUwgUAMIVwAQBMIVwAAFMIFwDAFMIFADCFcAEATCFcAABTCBcAwBTCBQAwhXABAEwhXAAAUwgXAMCUgPN6e+gkiUQiCoVCfo8xImRmZnpa98UXX3haN2HChLjXBAIBT8caZpdtvzo6Ojyt83pO8vPz415z9OjRITsWcKHC4bAyMjIGfJ13XAAAUwgXAMAUwgUAMIVwAQBMIVwAAFMIFwDAFMIFADCFcAEATCFcAABTCBcAwBTCBQAwhXABAEwhXAAAU8b4PQCSp7u729O6BQsWeFq3ZcuWuNdkZWV5OlZbW5undZs3b/a0rr6+Pu41//vf/zwda8OGDZ7Weblju9djAX7iHRcAwBTCBQAwJe5wbd++XTfffLMKCgoUCAS0adOmmNeXLl2qQCAQs82fPz9R8wIAUlzc4erp6dHMmTNVV1c34D7z58/X4cOHo9ubb775g4YEAOBbcX84o7KyUpWVlYPuEwwGlZeX53koAAAGkpSfcTU2NionJ0dXXXWVli9frmPHjg24b29vryKRSMwGAMBAEh6u+fPn67XXXlNDQ4P++Mc/qqmpSZWVlTp79my/+9fW1ioUCkW3wsLCRI8EABhBEv57XIsXL47+e/r06ZoxY4amTJmixsZGzZs375z9a2pqVF1dHX0ciUSIFwBgQEn/OHxxcbGys7PV2tra7+vBYFAZGRkxGwAAA0l6uA4dOqRjx455+q1+AAC+L+5vFZ44cSLm3VN7e7v27NmjrKwsZWVl6amnntLChQuVl5entrY2PfLII7r88stVUVGR0MEBAKkp7nDt2rVLN910U/Txtz+fWrJkidauXau9e/fq1VdfVXd3twoKClReXq5nnnlGwWAwcVMDAFJW3OGaO3eunHMDvv7+++//oIHgv507d3paN2HChARPYtucOXM8rbvxxhs9revr64t7zb///W9PxwL8xL0KAQCmEC4AgCmECwBgCuECAJhCuAAAphAuAIAphAsAYArhAgCYQrgAAKYQLgCAKYQLAGAK4QIAmEK4AACmxH13eAAXZvz48Z7WebnLu6RB/2rDQDZs2ODpWICfeMcFADCFcAEATCFcAABTCBcAwBTCBQAwhXABAEwhXAAAUwgXAMAUwgUAMIVwAQBMIVwAAFMIFwDAFMIFADCFu8MDSfL+++/7PQIwIvGOCwBgCuECAJhCuAAAphAuAIAphAsAYArhAgCYQrgAAKYQLgCAKYQLAGAK4QIAmEK4AACmEC4AgCncZBdIkoqKCr9HAEYk3nEBAEwhXAAAUwgXAMAUwgUAMIVwAQBMIVwAAFMIFwDAFMIFADCFcAEATCFcAABTCBcAwBTCBQAwhXABAEzh7vBAkhQXF/s9AjAi8Y4LAGAK4QIAmBJXuGpra3XttdcqPT1dOTk5WrBggVpaWmL2OXXqlKqqqnTppZfqkksu0cKFC9XV1ZXQoQEAqSuucDU1Namqqko7duzQtm3bdObMGZWXl6unpye6z0MPPaT33ntP77zzjpqamtTR0aHbbrst4YMDAFJTXB/O2Lp1a8zj+vp65eTkaPfu3ZozZ47C4bD+9re/af369frFL34hSVq3bp1+8pOfaMeOHfr5z3+euMkBACnpB/2MKxwOS5KysrIkSbt379aZM2dUVlYW3Wfq1KmaNGmSmpub+/0avb29ikQiMRsAAAPxHK6+vj6tXLlS1113naZNmyZJ6uzsVFpamjIzM2P2zc3NVWdnZ79fp7a2VqFQKLoVFhZ6HQkAkAI8h6uqqkr79u3Thg0bftAANTU1CofD0e3gwYM/6OsBAEY2T7+AvGLFCm3ZskXbt2/XxIkTo8/n5eXp9OnT6u7ujnnX1dXVpby8vH6/VjAYVDAY9DIGACAFxfWOyzmnFStWaOPGjfrwww9VVFQU8/qsWbM0duxYNTQ0RJ9raWnRgQMHVFpampiJAQApLa53XFVVVVq/fr02b96s9PT06M+tQqGQxo8fr1AopHvuuUfV1dXKyspSRkaGHnjgAZWWlvKJQgBAQsQVrrVr10qS5s6dG/P8unXrtHTpUknSn/70J40aNUoLFy5Ub2+vKioq9NJLLyVkWAAA4gqXc+68+4wbN051dXWqq6vzPBQAAAPh7vBAkvzjH//wtG7UKG8f9u3r6/O0DrCGm+wCAEwhXAAAUwgXAMAUwgUAMIVwAQBMIVwAAFMIFwDAFMIFADCFcAEATCFcAABTCBcAwBTCBQAwhZvsAkmyb98+T+v279/vaV1xcXHca6ZMmeLpWEePHvW0DkgE3nEBAEwhXAAAUwgXAMAUwgUAMIVwAQBMIVwAAFMIFwDAFMIFADCFcAEATCFcAABTCBcAwBTCBQAwhXABAEwJOOec30N8VyQSUSgU8nsMwDdLly71tO6VV16Je01TU5OnYz3wwAOe1n3++eee1iG1hMNhZWRkDPg677gAAKYQLgCAKYQLAGAK4QIAmEK4AACmEC4AgCmECwBgCuECAJhCuAAAphAuAIAphAsAYArhAgCYQrgAAKZwd3hgmBnsrtiDefvtt+NeU1ZW5ulY7777rqd1d999t6d1PT09ntbBJu4ODwAYUQgXAMAUwgUAMIVwAQBMIVwAAFMIFwDAFMIFADCFcAEATCFcAABTCBcAwBTCBQAwhXABAEwhXAAAU7g7PDBCeLmr/O9//3tPx1q+fLmndTNmzPC07vPPP/e0DjZxd3gAwIhCuAAApsQVrtraWl177bVKT09XTk6OFixYoJaWlph95s6dq0AgELPdd999CR0aAJC64gpXU1OTqqqqtGPHDm3btk1nzpxReXn5OX+d9N5779Xhw4ej25o1axI6NAAgdY2JZ+etW7fGPK6vr1dOTo52796tOXPmRJ+/6KKLlJeXl5gJAQD4jh/0M65wOCxJysrKinn+jTfeUHZ2tqZNm6aamhqdPHlywK/R29urSCQSswEAMJC43nF9V19fn1auXKnrrrtO06ZNiz5/xx13aPLkySooKNDevXv16KOPqqWlRe+++26/X6e2tlZPPfWU1zEAACnGc7iqqqq0b98+ffzxxzHPL1u2LPrv6dOnKz8/X/PmzVNbW5umTJlyztepqalRdXV19HEkElFhYaHXsQAAI5yncK1YsUJbtmzR9u3bNXHixEH3LSkpkSS1trb2G65gMKhgMOhlDABACoorXM45PfDAA9q4caMaGxtVVFR03jV79uyRJOXn53saEACA74orXFVVVVq/fr02b96s9PR0dXZ2SpJCoZDGjx+vtrY2rV+/Xr/85S916aWXau/evXrooYc0Z84cz7d6AQDgu+IK19q1ayV980vG37Vu3TotXbpUaWlp+uCDD/TCCy+op6dHhYWFWrhwoR5//PGEDQwASG3cZBcAMKxwk10AwIhCuAAAphAuAIAphAsAYArhAgCYQrgAAKYQLgCAKYQLAGAK4QIAmEK4AACmEC4AgCmECwBgyrAL1zC75y8AYIidrwPDLlzHjx/3ewQAgI/O14Fh92dN+vr61NHRofT0dAUCgZjXIpGICgsLdfDgwUFveZ9KOCexOB/n4pzE4nzEGk7nwzmn48ePq6CgQKNGDfy+Kq4/JDkURo0apYkTJw66T0ZGhu8neLjhnMTifJyLcxKL8xFruJyPC/l7jMPuW4UAAAyGcAEATDEVrmAwqNWrVysYDPo9yrDBOYnF+TgX5yQW5yOWxfMx7D6cAQDAYEy94wIAgHABAEwhXAAAUwgXAMAUU+Gqq6vTZZddpnHjxqmkpESffPKJ3yP54sknn1QgEIjZpk6d6vdYQ2r79u26+eabVVBQoEAgoE2bNsW87pzTqlWrlJ+fr/Hjx6usrEz79+/3Z9ghcL7zsXTp0nOumfnz5/sz7BCora3Vtddeq/T0dOXk5GjBggVqaWmJ2efUqVOqqqrSpZdeqksuuUQLFy5UV1eXTxMn34Wck7lz555zndx3330+TTwwM+F66623VF1drdWrV+vTTz/VzJkzVVFRoSNHjvg9mi+uvvpqHT58OLp9/PHHfo80pHp6ejRz5kzV1dX1+/qaNWv04osv6uWXX9bOnTt18cUXq6KiQqdOnRriSYfG+c6HJM2fPz/mmnnzzTeHcMKh1dTUpKqqKu3YsUPbtm3TmTNnVF5erp6enug+Dz30kN577z298847ampqUkdHh2677TYfp06uCzknknTvvffGXCdr1qzxaeJBOCNmz57tqqqqoo/Pnj3rCgoKXG1trY9T+WP16tVu5syZfo8xbEhyGzdujD7u6+tzeXl57rnnnos+193d7YLBoHvzzTd9mHBoff98OOfckiVL3C233OLLPMPBkSNHnCTX1NTknPvmehg7dqx75513ovt88cUXTpJrbm72a8wh9f1z4pxzN954o/v1r3/t31AXyMQ7rtOnT2v37t0qKyuLPjdq1CiVlZWpubnZx8n8s3//fhUUFKi4uFh33nmnDhw44PdIw0Z7e7s6OztjrpdQKKSSkpKUvV4kqbGxUTk5Obrqqqu0fPlyHTt2zO+Rhkw4HJYkZWVlSZJ2796tM2fOxFwjU6dO1aRJk1LmGvn+OfnWG2+8oezsbE2bNk01NTU6efKkH+MNatjdZLc/X331lc6ePavc3NyY53Nzc/Xll1/6NJV/SkpKVF9fr6uuukqHDx/WU089pRtuuEH79u1Tenq63+P5rrOzU5L6vV6+fS3VzJ8/X7fddpuKiorU1tamxx57TJWVlWpubtbo0aP9Hi+p+vr6tHLlSl133XWaNm2apG+ukbS0NGVmZsbsmyrXSH/nRJLuuOMOTZ48WQUFBdq7d68effRRtbS06N133/Vx2nOZCBdiVVZWRv89Y8YMlZSUaPLkyXr77bd1zz33+DgZhqvFixdH/z19+nTNmDFDU6ZMUWNjo+bNm+fjZMlXVVWlffv2pdzPgQcz0DlZtmxZ9N/Tp09Xfn6+5s2bp7a2Nk2ZMmWoxxyQiW8VZmdna/To0ed84qerq0t5eXk+TTV8ZGZm6sorr1Rra6vfowwL314TXC8DKy4uVnZ29oi/ZlasWKEtW7boo48+ivlzSXl5eTp9+rS6u7tj9k+Fa2Sgc9KfkpISSRp214mJcKWlpWnWrFlqaGiIPtfX16eGhgaVlpb6ONnwcOLECbW1tSk/P9/vUYaFoqIi5eXlxVwvkUhEO3fu5Hr5fw4dOqRjx46N2GvGOacVK1Zo48aN+vDDD1VUVBTz+qxZszR27NiYa6SlpUUHDhwYsdfI+c5Jf/bs2SNJw+868fvTIRdqw4YNLhgMuvr6evf555+7ZcuWuczMTNfZ2en3aEPuN7/5jWtsbHTt7e3un//8pysrK3PZ2dnuyJEjfo82ZI4fP+4+++wz99lnnzlJ7vnnn3efffaZ++9//+ucc+4Pf/iDy8zMdJs3b3Z79+51t9xyiysqKnJff/21z5Mnx2Dn4/jx4+7hhx92zc3Nrr293X3wwQfuZz/7mbviiivcqVOn/B49KZYvX+5CoZBrbGx0hw8fjm4nT56M7nPfffe5SZMmuQ8//NDt2rXLlZaWutLSUh+nTq7znZPW1lb39NNPu127drn29na3efNmV1xc7ObMmePz5OcyEy7nnPvLX/7iJk2a5NLS0tzs2bPdjh07/B7JF4sWLXL5+fkuLS3N/fjHP3aLFi1yra2tfo81pD766CMn6ZxtyZIlzrlvPhL/xBNPuNzcXBcMBt28efNcS0uLv0Mn0WDn4+TJk668vNxNmDDBjR071k2ePNnde++9I/p/+vo7F5LcunXrovt8/fXX7v7773c/+tGP3EUXXeRuvfVWd/jwYf+GTrLznZMDBw64OXPmuKysLBcMBt3ll1/ufvvb37pwOOzv4P3gz5oAAEwx8TMuAAC+RbgAAKYQLgCAKYQLAGAK4QIAmEK4AACmEC4AgCmECwBgCuECAJhCuAAAphAuAIAphAsAYMr/Ac55ZR2k942eAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(x_train[4, 1:], cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 14,
     "status": "ok",
     "timestamp": 1700970000164,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "Rg0WJuW5KLZ6",
    "outputId": "2a251f1f-f9e0-4096-a6f4-8b1608c45540"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 784)\n",
      "(10000, 784)\n"
     ]
    }
   ],
   "source": [
    "x_trainf=x_train.reshape(60000,784)\n",
    "x_testf=x_test.reshape(10000,784)\n",
    "print(x_trainf.shape)\n",
    "print(x_testf.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "executionInfo": {
     "elapsed": 9,
     "status": "ok",
     "timestamp": 1700970000164,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "00n-0shrKJjh"
   },
   "outputs": [],
   "source": [
    "x_trainf = x_trainf/255\n",
    "x_testf  = x_testf/255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "executionInfo": {
     "elapsed": 10,
     "status": "ok",
     "timestamp": 1700970000165,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "xEpxhRdCKJnJ"
   },
   "outputs": [],
   "source": [
    "model = tf.keras.Sequential(layers=[\n",
    "    tf.keras.layers.Dense(4, activation=tf.nn.relu),\n",
    "    tf.keras.layers.Dense(50, activation=tf.nn.relu),\n",
    "    tf.keras.layers.Dense(len(np.unique(y_train)), activation=tf.nn.softmax)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "executionInfo": {
     "elapsed": 9,
     "status": "ok",
     "timestamp": 1700970000165,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "TeDc7i_mrkGK"
   },
   "outputs": [],
   "source": [
    "model.compile(optimizer=tf.keras.optimizers.SGD(),\n",
    "              loss=tf.keras.losses.SparseCategoricalCrossentropy(),\n",
    "              metrics=tf.keras.metrics.SparseCategoricalAccuracy())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 622802,
     "status": "ok",
     "timestamp": 1700970622958,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "X7TVcta8KPUU",
    "outputId": "335fe4ba-cffe-4a0d-d930-8f06382bd045"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "1500/1500 [==============================] - 6s 3ms/step - loss: 1.8146 - sparse_categorical_accuracy: 0.3477 - val_loss: 1.1730 - val_sparse_categorical_accuracy: 0.6288\n",
      "Epoch 2/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.8676 - sparse_categorical_accuracy: 0.7208 - val_loss: 0.6658 - val_sparse_categorical_accuracy: 0.7931\n",
      "Epoch 3/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.6404 - sparse_categorical_accuracy: 0.8022 - val_loss: 0.5711 - val_sparse_categorical_accuracy: 0.8225\n",
      "Epoch 4/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.5792 - sparse_categorical_accuracy: 0.8230 - val_loss: 0.5289 - val_sparse_categorical_accuracy: 0.8388\n",
      "Epoch 5/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.5457 - sparse_categorical_accuracy: 0.8371 - val_loss: 0.5138 - val_sparse_categorical_accuracy: 0.8456\n",
      "Epoch 6/100\n",
      "1500/1500 [==============================] - 5s 4ms/step - loss: 0.5220 - sparse_categorical_accuracy: 0.8456 - val_loss: 0.4858 - val_sparse_categorical_accuracy: 0.8522\n",
      "Epoch 7/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.5054 - sparse_categorical_accuracy: 0.8518 - val_loss: 0.4665 - val_sparse_categorical_accuracy: 0.8591\n",
      "Epoch 8/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.4911 - sparse_categorical_accuracy: 0.8566 - val_loss: 0.4592 - val_sparse_categorical_accuracy: 0.8633\n",
      "Epoch 9/100\n",
      "1500/1500 [==============================] - 5s 4ms/step - loss: 0.4782 - sparse_categorical_accuracy: 0.8610 - val_loss: 0.4568 - val_sparse_categorical_accuracy: 0.8623\n",
      "Epoch 10/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.4675 - sparse_categorical_accuracy: 0.8640 - val_loss: 0.4434 - val_sparse_categorical_accuracy: 0.8687\n",
      "Epoch 11/100\n",
      "1500/1500 [==============================] - 7s 4ms/step - loss: 0.4567 - sparse_categorical_accuracy: 0.8669 - val_loss: 0.4385 - val_sparse_categorical_accuracy: 0.8693\n",
      "Epoch 12/100\n",
      "1500/1500 [==============================] - 5s 4ms/step - loss: 0.4491 - sparse_categorical_accuracy: 0.8689 - val_loss: 0.4289 - val_sparse_categorical_accuracy: 0.8725\n",
      "Epoch 13/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.4413 - sparse_categorical_accuracy: 0.8709 - val_loss: 0.4170 - val_sparse_categorical_accuracy: 0.8761\n",
      "Epoch 14/100\n",
      "1500/1500 [==============================] - 7s 5ms/step - loss: 0.4354 - sparse_categorical_accuracy: 0.8731 - val_loss: 0.4152 - val_sparse_categorical_accuracy: 0.8766\n",
      "Epoch 15/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.4298 - sparse_categorical_accuracy: 0.8751 - val_loss: 0.4257 - val_sparse_categorical_accuracy: 0.8720\n",
      "Epoch 16/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.4251 - sparse_categorical_accuracy: 0.8767 - val_loss: 0.4059 - val_sparse_categorical_accuracy: 0.8797\n",
      "Epoch 17/100\n",
      "1500/1500 [==============================] - 7s 5ms/step - loss: 0.4211 - sparse_categorical_accuracy: 0.8774 - val_loss: 0.4084 - val_sparse_categorical_accuracy: 0.8801\n",
      "Epoch 18/100\n",
      "1500/1500 [==============================] - 7s 5ms/step - loss: 0.4176 - sparse_categorical_accuracy: 0.8778 - val_loss: 0.4082 - val_sparse_categorical_accuracy: 0.8791\n",
      "Epoch 19/100\n",
      "1500/1500 [==============================] - 12s 8ms/step - loss: 0.4143 - sparse_categorical_accuracy: 0.8798 - val_loss: 0.3990 - val_sparse_categorical_accuracy: 0.8804\n",
      "Epoch 20/100\n",
      "1500/1500 [==============================] - 10s 7ms/step - loss: 0.4111 - sparse_categorical_accuracy: 0.8799 - val_loss: 0.4001 - val_sparse_categorical_accuracy: 0.8814\n",
      "Epoch 21/100\n",
      "1500/1500 [==============================] - 10s 7ms/step - loss: 0.4083 - sparse_categorical_accuracy: 0.8797 - val_loss: 0.4106 - val_sparse_categorical_accuracy: 0.8785\n",
      "Epoch 22/100\n",
      "1500/1500 [==============================] - 11s 7ms/step - loss: 0.4062 - sparse_categorical_accuracy: 0.8801 - val_loss: 0.3965 - val_sparse_categorical_accuracy: 0.8827\n",
      "Epoch 23/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.4036 - sparse_categorical_accuracy: 0.8817 - val_loss: 0.3951 - val_sparse_categorical_accuracy: 0.8813\n",
      "Epoch 24/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.4012 - sparse_categorical_accuracy: 0.8826 - val_loss: 0.4068 - val_sparse_categorical_accuracy: 0.8772\n",
      "Epoch 25/100\n",
      "1500/1500 [==============================] - 7s 4ms/step - loss: 0.3990 - sparse_categorical_accuracy: 0.8838 - val_loss: 0.3932 - val_sparse_categorical_accuracy: 0.8842\n",
      "Epoch 26/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3970 - sparse_categorical_accuracy: 0.8831 - val_loss: 0.3904 - val_sparse_categorical_accuracy: 0.8830\n",
      "Epoch 27/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3950 - sparse_categorical_accuracy: 0.8836 - val_loss: 0.3885 - val_sparse_categorical_accuracy: 0.8844\n",
      "Epoch 28/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3936 - sparse_categorical_accuracy: 0.8838 - val_loss: 0.3865 - val_sparse_categorical_accuracy: 0.8825\n",
      "Epoch 29/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3925 - sparse_categorical_accuracy: 0.8842 - val_loss: 0.3842 - val_sparse_categorical_accuracy: 0.8857\n",
      "Epoch 30/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3903 - sparse_categorical_accuracy: 0.8844 - val_loss: 0.3924 - val_sparse_categorical_accuracy: 0.8819\n",
      "Epoch 31/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3890 - sparse_categorical_accuracy: 0.8855 - val_loss: 0.3897 - val_sparse_categorical_accuracy: 0.8826\n",
      "Epoch 32/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3871 - sparse_categorical_accuracy: 0.8861 - val_loss: 0.3842 - val_sparse_categorical_accuracy: 0.8838\n",
      "Epoch 33/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3856 - sparse_categorical_accuracy: 0.8863 - val_loss: 0.3806 - val_sparse_categorical_accuracy: 0.8849\n",
      "Epoch 34/100\n",
      "1500/1500 [==============================] - 7s 4ms/step - loss: 0.3845 - sparse_categorical_accuracy: 0.8870 - val_loss: 0.3902 - val_sparse_categorical_accuracy: 0.8824\n",
      "Epoch 35/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3843 - sparse_categorical_accuracy: 0.8861 - val_loss: 0.3825 - val_sparse_categorical_accuracy: 0.8818\n",
      "Epoch 36/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3824 - sparse_categorical_accuracy: 0.8878 - val_loss: 0.3785 - val_sparse_categorical_accuracy: 0.8865\n",
      "Epoch 37/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3813 - sparse_categorical_accuracy: 0.8875 - val_loss: 0.3871 - val_sparse_categorical_accuracy: 0.8835\n",
      "Epoch 38/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3801 - sparse_categorical_accuracy: 0.8878 - val_loss: 0.3813 - val_sparse_categorical_accuracy: 0.8853\n",
      "Epoch 39/100\n",
      "1500/1500 [==============================] - 8s 5ms/step - loss: 0.3787 - sparse_categorical_accuracy: 0.8881 - val_loss: 0.3976 - val_sparse_categorical_accuracy: 0.8803\n",
      "Epoch 40/100\n",
      "1500/1500 [==============================] - 5s 4ms/step - loss: 0.3785 - sparse_categorical_accuracy: 0.8882 - val_loss: 0.3815 - val_sparse_categorical_accuracy: 0.8846\n",
      "Epoch 41/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3785 - sparse_categorical_accuracy: 0.8881 - val_loss: 0.3744 - val_sparse_categorical_accuracy: 0.8860\n",
      "Epoch 42/100\n",
      "1500/1500 [==============================] - 7s 5ms/step - loss: 0.3764 - sparse_categorical_accuracy: 0.8890 - val_loss: 0.3799 - val_sparse_categorical_accuracy: 0.8843\n",
      "Epoch 43/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3760 - sparse_categorical_accuracy: 0.8884 - val_loss: 0.3745 - val_sparse_categorical_accuracy: 0.8877\n",
      "Epoch 44/100\n",
      "1500/1500 [==============================] - 5s 4ms/step - loss: 0.3756 - sparse_categorical_accuracy: 0.8895 - val_loss: 0.3820 - val_sparse_categorical_accuracy: 0.8821\n",
      "Epoch 45/100\n",
      "1500/1500 [==============================] - 7s 5ms/step - loss: 0.3748 - sparse_categorical_accuracy: 0.8881 - val_loss: 0.3805 - val_sparse_categorical_accuracy: 0.8824\n",
      "Epoch 46/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3728 - sparse_categorical_accuracy: 0.8885 - val_loss: 0.3723 - val_sparse_categorical_accuracy: 0.8859\n",
      "Epoch 47/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3726 - sparse_categorical_accuracy: 0.8892 - val_loss: 0.3774 - val_sparse_categorical_accuracy: 0.8848\n",
      "Epoch 48/100\n",
      "1500/1500 [==============================] - 7s 5ms/step - loss: 0.3721 - sparse_categorical_accuracy: 0.8894 - val_loss: 0.3799 - val_sparse_categorical_accuracy: 0.8855\n",
      "Epoch 49/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3714 - sparse_categorical_accuracy: 0.8897 - val_loss: 0.3751 - val_sparse_categorical_accuracy: 0.8849\n",
      "Epoch 50/100\n",
      "1500/1500 [==============================] - 7s 4ms/step - loss: 0.3702 - sparse_categorical_accuracy: 0.8905 - val_loss: 0.3780 - val_sparse_categorical_accuracy: 0.8846\n",
      "Epoch 51/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3695 - sparse_categorical_accuracy: 0.8914 - val_loss: 0.3765 - val_sparse_categorical_accuracy: 0.8842\n",
      "Epoch 52/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3694 - sparse_categorical_accuracy: 0.8896 - val_loss: 0.4004 - val_sparse_categorical_accuracy: 0.8766\n",
      "Epoch 53/100\n",
      "1500/1500 [==============================] - 7s 5ms/step - loss: 0.3679 - sparse_categorical_accuracy: 0.8912 - val_loss: 0.3733 - val_sparse_categorical_accuracy: 0.8860\n",
      "Epoch 54/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3676 - sparse_categorical_accuracy: 0.8898 - val_loss: 0.3720 - val_sparse_categorical_accuracy: 0.8868\n",
      "Epoch 55/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3671 - sparse_categorical_accuracy: 0.8913 - val_loss: 0.3719 - val_sparse_categorical_accuracy: 0.8851\n",
      "Epoch 56/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3668 - sparse_categorical_accuracy: 0.8918 - val_loss: 0.3655 - val_sparse_categorical_accuracy: 0.8878\n",
      "Epoch 57/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3656 - sparse_categorical_accuracy: 0.8915 - val_loss: 0.3710 - val_sparse_categorical_accuracy: 0.8876\n",
      "Epoch 58/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3655 - sparse_categorical_accuracy: 0.8916 - val_loss: 0.3683 - val_sparse_categorical_accuracy: 0.8876\n",
      "Epoch 59/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3652 - sparse_categorical_accuracy: 0.8917 - val_loss: 0.3695 - val_sparse_categorical_accuracy: 0.8863\n",
      "Epoch 60/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3651 - sparse_categorical_accuracy: 0.8901 - val_loss: 0.3648 - val_sparse_categorical_accuracy: 0.8890\n",
      "Epoch 61/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3641 - sparse_categorical_accuracy: 0.8912 - val_loss: 0.3719 - val_sparse_categorical_accuracy: 0.8864\n",
      "Epoch 62/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3633 - sparse_categorical_accuracy: 0.8909 - val_loss: 0.3652 - val_sparse_categorical_accuracy: 0.8880\n",
      "Epoch 63/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3629 - sparse_categorical_accuracy: 0.8917 - val_loss: 0.3644 - val_sparse_categorical_accuracy: 0.8889\n",
      "Epoch 64/100\n",
      "1500/1500 [==============================] - 7s 4ms/step - loss: 0.3627 - sparse_categorical_accuracy: 0.8914 - val_loss: 0.3666 - val_sparse_categorical_accuracy: 0.8877\n",
      "Epoch 65/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3621 - sparse_categorical_accuracy: 0.8923 - val_loss: 0.3631 - val_sparse_categorical_accuracy: 0.8882\n",
      "Epoch 66/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3616 - sparse_categorical_accuracy: 0.8921 - val_loss: 0.3674 - val_sparse_categorical_accuracy: 0.8878\n",
      "Epoch 67/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3614 - sparse_categorical_accuracy: 0.8926 - val_loss: 0.3668 - val_sparse_categorical_accuracy: 0.8873\n",
      "Epoch 68/100\n",
      "1500/1500 [==============================] - 5s 4ms/step - loss: 0.3605 - sparse_categorical_accuracy: 0.8927 - val_loss: 0.3698 - val_sparse_categorical_accuracy: 0.8852\n",
      "Epoch 69/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3596 - sparse_categorical_accuracy: 0.8925 - val_loss: 0.3638 - val_sparse_categorical_accuracy: 0.8906\n",
      "Epoch 70/100\n",
      "1500/1500 [==============================] - 7s 5ms/step - loss: 0.3597 - sparse_categorical_accuracy: 0.8923 - val_loss: 0.3741 - val_sparse_categorical_accuracy: 0.8842\n",
      "Epoch 71/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3588 - sparse_categorical_accuracy: 0.8916 - val_loss: 0.3708 - val_sparse_categorical_accuracy: 0.8867\n",
      "Epoch 72/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3585 - sparse_categorical_accuracy: 0.8930 - val_loss: 0.3658 - val_sparse_categorical_accuracy: 0.8878\n",
      "Epoch 73/100\n",
      "1500/1500 [==============================] - 7s 4ms/step - loss: 0.3585 - sparse_categorical_accuracy: 0.8928 - val_loss: 0.3665 - val_sparse_categorical_accuracy: 0.8879\n",
      "Epoch 74/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3577 - sparse_categorical_accuracy: 0.8929 - val_loss: 0.3799 - val_sparse_categorical_accuracy: 0.8841\n",
      "Epoch 75/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3582 - sparse_categorical_accuracy: 0.8937 - val_loss: 0.3690 - val_sparse_categorical_accuracy: 0.8868\n",
      "Epoch 76/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3572 - sparse_categorical_accuracy: 0.8920 - val_loss: 0.3724 - val_sparse_categorical_accuracy: 0.8852\n",
      "Epoch 77/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3565 - sparse_categorical_accuracy: 0.8932 - val_loss: 0.3712 - val_sparse_categorical_accuracy: 0.8848\n",
      "Epoch 78/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3565 - sparse_categorical_accuracy: 0.8935 - val_loss: 0.3672 - val_sparse_categorical_accuracy: 0.8882\n",
      "Epoch 79/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3561 - sparse_categorical_accuracy: 0.8943 - val_loss: 0.3654 - val_sparse_categorical_accuracy: 0.8882\n",
      "Epoch 80/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3560 - sparse_categorical_accuracy: 0.8927 - val_loss: 0.3627 - val_sparse_categorical_accuracy: 0.8862\n",
      "Epoch 81/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3559 - sparse_categorical_accuracy: 0.8935 - val_loss: 0.3633 - val_sparse_categorical_accuracy: 0.8896\n",
      "Epoch 82/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3553 - sparse_categorical_accuracy: 0.8930 - val_loss: 0.3613 - val_sparse_categorical_accuracy: 0.8886\n",
      "Epoch 83/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3548 - sparse_categorical_accuracy: 0.8936 - val_loss: 0.3732 - val_sparse_categorical_accuracy: 0.8850\n",
      "Epoch 84/100\n",
      "1500/1500 [==============================] - 7s 4ms/step - loss: 0.3541 - sparse_categorical_accuracy: 0.8939 - val_loss: 0.3650 - val_sparse_categorical_accuracy: 0.8869\n",
      "Epoch 85/100\n",
      "1500/1500 [==============================] - 5s 4ms/step - loss: 0.3534 - sparse_categorical_accuracy: 0.8944 - val_loss: 0.3631 - val_sparse_categorical_accuracy: 0.8875\n",
      "Epoch 86/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3532 - sparse_categorical_accuracy: 0.8942 - val_loss: 0.3599 - val_sparse_categorical_accuracy: 0.8884\n",
      "Epoch 87/100\n",
      "1500/1500 [==============================] - 7s 5ms/step - loss: 0.3523 - sparse_categorical_accuracy: 0.8944 - val_loss: 0.3618 - val_sparse_categorical_accuracy: 0.8890\n",
      "Epoch 88/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3533 - sparse_categorical_accuracy: 0.8939 - val_loss: 0.3627 - val_sparse_categorical_accuracy: 0.8887\n",
      "Epoch 89/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3524 - sparse_categorical_accuracy: 0.8944 - val_loss: 0.3566 - val_sparse_categorical_accuracy: 0.8901\n",
      "Epoch 90/100\n",
      "1500/1500 [==============================] - 7s 5ms/step - loss: 0.3520 - sparse_categorical_accuracy: 0.8942 - val_loss: 0.3587 - val_sparse_categorical_accuracy: 0.8903\n",
      "Epoch 91/100\n",
      "1500/1500 [==============================] - 5s 4ms/step - loss: 0.3520 - sparse_categorical_accuracy: 0.8946 - val_loss: 0.3633 - val_sparse_categorical_accuracy: 0.8881\n",
      "Epoch 92/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3513 - sparse_categorical_accuracy: 0.8946 - val_loss: 0.3587 - val_sparse_categorical_accuracy: 0.8924\n",
      "Epoch 93/100\n",
      "1500/1500 [==============================] - 7s 4ms/step - loss: 0.3516 - sparse_categorical_accuracy: 0.8944 - val_loss: 0.3710 - val_sparse_categorical_accuracy: 0.8839\n",
      "Epoch 94/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3512 - sparse_categorical_accuracy: 0.8941 - val_loss: 0.3594 - val_sparse_categorical_accuracy: 0.8883\n",
      "Epoch 95/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3502 - sparse_categorical_accuracy: 0.8951 - val_loss: 0.3630 - val_sparse_categorical_accuracy: 0.8887\n",
      "Epoch 96/100\n",
      "1500/1500 [==============================] - 7s 4ms/step - loss: 0.3496 - sparse_categorical_accuracy: 0.8946 - val_loss: 0.3598 - val_sparse_categorical_accuracy: 0.8898\n",
      "Epoch 97/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3496 - sparse_categorical_accuracy: 0.8950 - val_loss: 0.3696 - val_sparse_categorical_accuracy: 0.8856\n",
      "Epoch 98/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3493 - sparse_categorical_accuracy: 0.8950 - val_loss: 0.3585 - val_sparse_categorical_accuracy: 0.8913\n",
      "Epoch 99/100\n",
      "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3494 - sparse_categorical_accuracy: 0.8950 - val_loss: 0.3667 - val_sparse_categorical_accuracy: 0.8883\n",
      "Epoch 100/100\n",
      "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3490 - sparse_categorical_accuracy: 0.8951 - val_loss: 0.3588 - val_sparse_categorical_accuracy: 0.8901\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(x_trainf, y_train, epochs=100, validation_split=0.2, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 25,
     "status": "ok",
     "timestamp": 1700970622959,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "GCiu_P9cKJqA",
    "outputId": "94a28d90-0a44-46b8-b9ed-683c53b2e2d4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense_3 (Dense)             (32, 4)                   3140      \n",
      "                                                                 \n",
      " dense_4 (Dense)             (32, 50)                  250       \n",
      "                                                                 \n",
      " dense_5 (Dense)             (32, 10)                  510       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 3900 (15.23 KB)\n",
      "Trainable params: 3900 (15.23 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1024,
     "status": "ok",
     "timestamp": 1700970623968,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "ji8GyHC5KJsv",
    "outputId": "7c69a877-3bc7-45dd-c4cd-359beb0fce8a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 1s 3ms/step - loss: 0.3737 - sparse_categorical_accuracy: 0.8856\n",
      "Test set loss: 0.37, test set accuracy: 88.56%\n"
     ]
    }
   ],
   "source": [
    "results = model.evaluate(x_testf, y_test)\n",
    "print(f'Test set loss: {results[0]:0.2f}, test set accuracy: {results[1]*100:0.2f}%')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 504
    },
    "executionInfo": {
     "elapsed": 1099,
     "status": "ok",
     "timestamp": 1700970625056,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "3Rfl1h7_jP0x",
    "outputId": "6dbb0115-33eb-40de-b8b0-48aa5c64bbb7"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Loss Curves')"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1300x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig3, ax3 = plt.subplots(1, 2, figsize=(13, 5))\n",
    "\n",
    "ax3[0].plot(history.history['sparse_categorical_accuracy'])\n",
    "ax3[0].plot(history.history['val_sparse_categorical_accuracy'])\n",
    "ax3[0].set_xlabel('Epoch')\n",
    "ax3[0].set_ylabel('Accuracy')\n",
    "ax3[0].legend(['Training Accuracy', 'Validation Accuracy'], loc='lower right')\n",
    "ax3[0].set_title('Accuracy Curves')\n",
    "\n",
    "ax3[1].plot(history.history['loss'])\n",
    "ax3[1].plot(history.history['val_loss'])\n",
    "\n",
    "ax3[1].set_xlabel('Epoch')\n",
    "ax3[1].set_ylabel('Loss')\n",
    "ax3[1].legend(['Training Loss', 'Validation Loss'])\n",
    "ax3[1].set_title('Loss Curves')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "mUWSQYigpbij"
   },
   "source": [
    "### Convolutional Neural Network for Digit Recognition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 31,
     "status": "ok",
     "timestamp": 1700970625056,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "x4bG1kkcKJye",
    "outputId": "ddcf593c-059c-4b26-dcc8-5a1bdefd27b6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 28, 28)\n",
      "(10000, 28, 28)\n"
     ]
    }
   ],
   "source": [
    "(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()\n",
    "x_train = np.expand_dims(x_train, -1)\n",
    "x_test = np.expand_dims(x_test, -1)\n",
    "print(x_train.shape)\n",
    "print(x_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "executionInfo": {
     "elapsed": 26,
     "status": "ok",
     "timestamp": 1700970625057,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "yI5LTBHbKJ4P"
   },
   "outputs": [],
   "source": [
    "x_train = x_train.astype(\"float32\") / 255\n",
    "x_test = x_test.astype(\"float32\") / 255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 25,
     "status": "ok",
     "timestamp": 1700970625057,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "-HJjQsGAKJ9Y",
    "outputId": "f1ea01cd-04b1-46ff-a17b-507f79d3b606"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of Classes: 10\n"
     ]
    }
   ],
   "source": [
    "number_of_classes = len(np.unique(y_train))\n",
    "print(f'Number of Classes: {number_of_classes}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 19,
     "status": "ok",
     "timestamp": 1700970625058,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "a0eHTB0dJqkL",
    "outputId": "56264602-59ca-4fa0-c79c-5f211bda580f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d (Conv2D)             (None, 26, 26, 32)        320       \n",
      "                                                                 \n",
      " max_pooling2d (MaxPooling2  (None, 13, 13, 32)        0         \n",
      " D)                                                              \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 11, 11, 64)        18496     \n",
      "                                                                 \n",
      " max_pooling2d_1 (MaxPoolin  (None, 5, 5, 64)          0         \n",
      " g2D)                                                            \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 1600)              0         \n",
      "                                                                 \n",
      " dropout (Dropout)           (None, 1600)              0         \n",
      "                                                                 \n",
      " dense_6 (Dense)             (None, 10)                16010     \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 34826 (136.04 KB)\n",
      "Trainable params: 34826 (136.04 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = keras.Sequential([ keras.Input(shape=(28, 28, 1)),\n",
    "                          layers.Conv2D(32, kernel_size=(3, 3), activation=\"relu\"),\n",
    "                           layers.MaxPooling2D(pool_size=(2, 2)),\n",
    "                           layers.Conv2D(64, kernel_size=(3, 3), activation=\"relu\"),\n",
    "                           layers.MaxPooling2D(pool_size=(2, 2)),\n",
    "                           layers.Flatten(),\n",
    "                           layers.Dropout(0.5),\n",
    "                           layers.Dense(number_of_classes, activation=\"softmax\"),\n",
    "                           ])\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "executionInfo": {
     "elapsed": 9,
     "status": "ok",
     "timestamp": 1700970625058,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "OgwbbfL1JqmZ"
   },
   "outputs": [],
   "source": [
    "y_train = keras.utils.to_categorical(y_train, number_of_classes)\n",
    "y_test = keras.utils.to_categorical(y_test, number_of_classes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "executionInfo": {
     "elapsed": 10,
     "status": "ok",
     "timestamp": 1700970625059,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "Pq1f5txoJqoq"
   },
   "outputs": [],
   "source": [
    "model.compile(loss=\"categorical_crossentropy\", optimizer=\"adam\", metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 42001,
     "status": "ok",
     "timestamp": 1700970667051,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "aTLvbCMSJqrM",
    "outputId": "1b36bdac-ee7f-43ca-ce78-bd647fc53304"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/15\n",
      "375/375 [==============================] - 6s 7ms/step - loss: 0.3850 - accuracy: 0.8834 - val_loss: 0.0983 - val_accuracy: 0.9727\n",
      "Epoch 2/15\n",
      "375/375 [==============================] - 3s 7ms/step - loss: 0.1167 - accuracy: 0.9647 - val_loss: 0.0668 - val_accuracy: 0.9811\n",
      "Epoch 3/15\n",
      "375/375 [==============================] - 3s 7ms/step - loss: 0.0883 - accuracy: 0.9726 - val_loss: 0.0547 - val_accuracy: 0.9844\n",
      "Epoch 4/15\n",
      "375/375 [==============================] - 2s 5ms/step - loss: 0.0733 - accuracy: 0.9769 - val_loss: 0.0487 - val_accuracy: 0.9862\n",
      "Epoch 5/15\n",
      "375/375 [==============================] - 2s 5ms/step - loss: 0.0642 - accuracy: 0.9802 - val_loss: 0.0450 - val_accuracy: 0.9868\n",
      "Epoch 6/15\n",
      "375/375 [==============================] - 2s 5ms/step - loss: 0.0558 - accuracy: 0.9827 - val_loss: 0.0434 - val_accuracy: 0.9878\n",
      "Epoch 7/15\n",
      "375/375 [==============================] - 2s 5ms/step - loss: 0.0529 - accuracy: 0.9836 - val_loss: 0.0375 - val_accuracy: 0.9882\n",
      "Epoch 8/15\n",
      "375/375 [==============================] - 2s 5ms/step - loss: 0.0487 - accuracy: 0.9848 - val_loss: 0.0370 - val_accuracy: 0.9898\n",
      "Epoch 9/15\n",
      "375/375 [==============================] - 2s 6ms/step - loss: 0.0456 - accuracy: 0.9854 - val_loss: 0.0362 - val_accuracy: 0.9893\n",
      "Epoch 10/15\n",
      "375/375 [==============================] - 2s 6ms/step - loss: 0.0439 - accuracy: 0.9859 - val_loss: 0.0360 - val_accuracy: 0.9899\n",
      "Epoch 11/15\n",
      "375/375 [==============================] - 3s 7ms/step - loss: 0.0399 - accuracy: 0.9873 - val_loss: 0.0397 - val_accuracy: 0.9877\n",
      "Epoch 12/15\n",
      "375/375 [==============================] - 2s 5ms/step - loss: 0.0374 - accuracy: 0.9879 - val_loss: 0.0353 - val_accuracy: 0.9893\n",
      "Epoch 13/15\n",
      "375/375 [==============================] - 2s 5ms/step - loss: 0.0381 - accuracy: 0.9873 - val_loss: 0.0328 - val_accuracy: 0.9896\n",
      "Epoch 14/15\n",
      "375/375 [==============================] - 2s 5ms/step - loss: 0.0357 - accuracy: 0.9882 - val_loss: 0.0311 - val_accuracy: 0.9911\n",
      "Epoch 15/15\n",
      "375/375 [==============================] - 2s 5ms/step - loss: 0.0335 - accuracy: 0.9892 - val_loss: 0.0330 - val_accuracy: 0.9901\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(x_train, y_train, batch_size=128, epochs=15, validation_split=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1419,
     "status": "ok",
     "timestamp": 1700970668456,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "clPq2G5xJqte",
    "outputId": "ff640dde-1190-46bc-9a20-39b839bb2aee"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test accuracy: 99.1200%\n"
     ]
    }
   ],
   "source": [
    "score = model.evaluate(x_test, y_test, verbose=0)\n",
    "print(f'Test accuracy: {score[1]*100:0.4f}%')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 504
    },
    "executionInfo": {
     "elapsed": 789,
     "status": "ok",
     "timestamp": 1700970669238,
     "user": {
      "displayName": "Arvin Singh",
      "userId": "10893380314688769536"
     },
     "user_tz": 0
    },
    "id": "MfG7lqSEjqM9",
    "outputId": "14c43f74-894e-4c7a-edf6-bda272a9f662"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Loss Curves')"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1300x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig4, ax4 = plt.subplots(1, 2, figsize=(13, 5))\n",
    "\n",
    "ax4[0].plot(history.history['accuracy'])\n",
    "ax4[0].plot(history.history['val_accuracy'])\n",
    "ax4[0].set_xlabel('Epoch')\n",
    "ax4[0].set_ylabel('Accuracy')\n",
    "ax4[0].legend(['Training Accuracy', 'Validation Accuracy'], loc='lower right')\n",
    "ax4[0].set_title('Accuracy Curves')\n",
    "\n",
    "ax4[1].plot(history.history['loss'])\n",
    "ax4[1].plot(history.history['val_loss'])\n",
    "\n",
    "ax4[1].set_xlabel('Epoch')\n",
    "ax4[1].set_ylabel('Loss')\n",
    "ax4[1].legend(['Training Loss', 'Validation Loss'])\n",
    "ax4[1].set_title('Loss Curves')"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "authorship_tag": "ABX9TyOfyyOv1kS+b65vIPpg1s9O",
   "gpuType": "T4",
   "provenance": []
  },
  "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}