HouseingPrice_Boston.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyM81IkoVKjmEMzDvaXwZPGq",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/akshaypat444/tensorflow_practice/blob/main/HouseingPrice_Boston.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "Chd-qP5tKyZc"
      },
      "outputs": [],
      "source": [
        "import tensorflow as tf\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.model_selection import train_test_split\n",
        "import seaborn as sns\n",
        "%matplotlib inline"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "(x_train, y_train), (x_test, y_test)  = tf.keras.datasets.boston_housing.load_data(\n",
        "    path='boston_housing.npz', test_split=0.2, seed=113\n",
        ")"
      ],
      "metadata": {
        "id": "uIbST8GiK6n7"
      },
      "execution_count": 49,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "x_train.shape, y_train.shape, x_test.shape, y_test.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GFM-Mp93wgZp",
        "outputId": "689b2e1a-795a-4b20-92e9-3532653b24c1"
      },
      "execution_count": 50,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((404, 13), (404,), (102, 13), (102,))"
            ]
          },
          "metadata": {},
          "execution_count": 50
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Model 1 - Without pre-processing"
      ],
      "metadata": {
        "id": "haYwEe_4w9wY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "tf.random.set_seed(42)\n",
        "\n",
        "# Create a Model\n",
        "boston_regression_model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(20),\n",
        "    tf.keras.layers.Dense(10),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# Compile a Model\n",
        "boston_regression_model.compile(loss=tf.keras.losses.mae,\n",
        "                                optimizer=tf.keras.optimizers.SGD(learning_rate=0.0001),\n",
        "                                metrics=[\"mae\"])\n",
        "\n",
        "# Fit a model\n",
        "boston_regression_model_hist = boston_regression_model.fit(x_train, y_train, epochs=150)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yTs3lsjuw8Dk",
        "outputId": "a13e186e-7859-48b7-f0fc-91d985d4ac62"
      },
      "execution_count": 87,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/150\n",
            "13/13 [==============================] - 1s 8ms/step - loss: 61.7529 - mae: 61.7529\n",
            "Epoch 2/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 55.3145 - mae: 55.3145\n",
            "Epoch 3/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 45.3927 - mae: 45.3927\n",
            "Epoch 4/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 40.9839 - mae: 40.9839\n",
            "Epoch 5/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 33.3632 - mae: 33.3632\n",
            "Epoch 6/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 30.3255 - mae: 30.3255\n",
            "Epoch 7/150\n",
            "13/13 [==============================] - 0s 4ms/step - loss: 25.4855 - mae: 25.4855\n",
            "Epoch 8/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 23.9401 - mae: 23.9401\n",
            "Epoch 9/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 19.4510 - mae: 19.4510\n",
            "Epoch 10/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 18.4416 - mae: 18.4416\n",
            "Epoch 11/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 14.9625 - mae: 14.9625\n",
            "Epoch 12/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 14.9311 - mae: 14.9311\n",
            "Epoch 13/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 15.4383 - mae: 15.4383\n",
            "Epoch 14/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 13.8645 - mae: 13.8645\n",
            "Epoch 15/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 11.1504 - mae: 11.1504\n",
            "Epoch 16/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 10.9214 - mae: 10.9214\n",
            "Epoch 17/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 10.7579 - mae: 10.7579\n",
            "Epoch 18/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 10.6298 - mae: 10.6298\n",
            "Epoch 19/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 9.9606 - mae: 9.9606\n",
            "Epoch 20/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 10.0745 - mae: 10.0745\n",
            "Epoch 21/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 7.7866 - mae: 7.7866\n",
            "Epoch 22/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 8.9950 - mae: 8.9950\n",
            "Epoch 23/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 9.1210 - mae: 9.1210\n",
            "Epoch 24/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 8.7085 - mae: 8.7085\n",
            "Epoch 25/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 7.7924 - mae: 7.7924\n",
            "Epoch 26/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 7.9588 - mae: 7.9588\n",
            "Epoch 27/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 8.3383 - mae: 8.3383\n",
            "Epoch 28/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 8.3835 - mae: 8.3835\n",
            "Epoch 29/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 7.9981 - mae: 7.9981\n",
            "Epoch 30/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 7.6961 - mae: 7.6961\n",
            "Epoch 31/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 7.0824 - mae: 7.0824\n",
            "Epoch 32/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.5456 - mae: 6.5456\n",
            "Epoch 33/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 7.5078 - mae: 7.5078\n",
            "Epoch 34/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.8211 - mae: 6.8211\n",
            "Epoch 35/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 7.4099 - mae: 7.4099\n",
            "Epoch 36/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 7.1461 - mae: 7.1461\n",
            "Epoch 37/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 7.1224 - mae: 7.1224\n",
            "Epoch 38/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.9531 - mae: 6.9531\n",
            "Epoch 39/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.9233 - mae: 6.9233\n",
            "Epoch 40/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.2768 - mae: 6.2768\n",
            "Epoch 41/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.6985 - mae: 6.6985\n",
            "Epoch 42/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.3089 - mae: 6.3089\n",
            "Epoch 43/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.5092 - mae: 6.5092\n",
            "Epoch 44/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.8696 - mae: 6.8696\n",
            "Epoch 45/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.7194 - mae: 6.7194\n",
            "Epoch 46/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.6924 - mae: 6.6924\n",
            "Epoch 47/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.7459 - mae: 6.7459\n",
            "Epoch 48/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1279 - mae: 6.1279\n",
            "Epoch 49/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0584 - mae: 6.0584\n",
            "Epoch 50/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.7161 - mae: 6.7161\n",
            "Epoch 51/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8601 - mae: 5.8601\n",
            "Epoch 52/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.5560 - mae: 6.5560\n",
            "Epoch 53/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.2669 - mae: 6.2669\n",
            "Epoch 54/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.8791 - mae: 6.8791\n",
            "Epoch 55/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.2554 - mae: 6.2554\n",
            "Epoch 56/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.3608 - mae: 6.3608\n",
            "Epoch 57/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.2797 - mae: 6.2797\n",
            "Epoch 58/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9833 - mae: 5.9833\n",
            "Epoch 59/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0154 - mae: 6.0154\n",
            "Epoch 60/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1737 - mae: 6.1737\n",
            "Epoch 61/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.2580 - mae: 6.2580\n",
            "Epoch 62/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8453 - mae: 5.8453\n",
            "Epoch 63/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.5169 - mae: 6.5169\n",
            "Epoch 64/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1805 - mae: 6.1805\n",
            "Epoch 65/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.7682 - mae: 6.7682\n",
            "Epoch 66/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1475 - mae: 6.1475\n",
            "Epoch 67/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.5598 - mae: 6.5598\n",
            "Epoch 68/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.5053 - mae: 6.5053\n",
            "Epoch 69/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.2290 - mae: 6.2290\n",
            "Epoch 70/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1103 - mae: 6.1103\n",
            "Epoch 71/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9127 - mae: 5.9127\n",
            "Epoch 72/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.5480 - mae: 6.5480\n",
            "Epoch 73/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0404 - mae: 6.0404\n",
            "Epoch 74/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9895 - mae: 5.9895\n",
            "Epoch 75/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.2682 - mae: 6.2682\n",
            "Epoch 76/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9525 - mae: 5.9525\n",
            "Epoch 77/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1978 - mae: 6.1978\n",
            "Epoch 78/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9613 - mae: 5.9613\n",
            "Epoch 79/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.7815 - mae: 5.7815\n",
            "Epoch 80/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1214 - mae: 6.1214\n",
            "Epoch 81/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.5253 - mae: 6.5253\n",
            "Epoch 82/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8060 - mae: 5.8060\n",
            "Epoch 83/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1699 - mae: 6.1699\n",
            "Epoch 84/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1265 - mae: 6.1265\n",
            "Epoch 85/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0385 - mae: 6.0385\n",
            "Epoch 86/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0312 - mae: 6.0312\n",
            "Epoch 87/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8829 - mae: 5.8829\n",
            "Epoch 88/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.8905 - mae: 5.8905\n",
            "Epoch 89/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9732 - mae: 5.9732\n",
            "Epoch 90/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1061 - mae: 6.1061\n",
            "Epoch 91/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9324 - mae: 5.9324\n",
            "Epoch 92/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0704 - mae: 6.0704\n",
            "Epoch 93/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0209 - mae: 6.0209\n",
            "Epoch 94/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1747 - mae: 6.1747\n",
            "Epoch 95/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0927 - mae: 6.0927\n",
            "Epoch 96/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1188 - mae: 6.1188\n",
            "Epoch 97/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9584 - mae: 5.9584\n",
            "Epoch 98/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8514 - mae: 5.8514\n",
            "Epoch 99/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0182 - mae: 6.0182\n",
            "Epoch 100/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.8677 - mae: 5.8677\n",
            "Epoch 101/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.1117 - mae: 6.1117\n",
            "Epoch 102/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8295 - mae: 5.8295\n",
            "Epoch 103/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8594 - mae: 5.8594\n",
            "Epoch 104/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.0617 - mae: 6.0617\n",
            "Epoch 105/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9895 - mae: 5.9895\n",
            "Epoch 106/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9722 - mae: 5.9722\n",
            "Epoch 107/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0983 - mae: 6.0983\n",
            "Epoch 108/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8800 - mae: 5.8800\n",
            "Epoch 109/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.8721 - mae: 5.8721\n",
            "Epoch 110/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8412 - mae: 5.8412\n",
            "Epoch 111/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.2524 - mae: 6.2524\n",
            "Epoch 112/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.7576 - mae: 5.7576\n",
            "Epoch 113/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.7232 - mae: 5.7232\n",
            "Epoch 114/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.7079 - mae: 5.7079\n",
            "Epoch 115/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1032 - mae: 6.1032\n",
            "Epoch 116/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9434 - mae: 5.9434\n",
            "Epoch 117/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0385 - mae: 6.0385\n",
            "Epoch 118/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8606 - mae: 5.8606\n",
            "Epoch 119/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9942 - mae: 5.9942\n",
            "Epoch 120/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0788 - mae: 6.0788\n",
            "Epoch 121/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8743 - mae: 5.8743\n",
            "Epoch 122/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.2056 - mae: 6.2056\n",
            "Epoch 123/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.8612 - mae: 5.8612\n",
            "Epoch 124/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0365 - mae: 6.0365\n",
            "Epoch 125/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9855 - mae: 5.9855\n",
            "Epoch 126/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.7919 - mae: 5.7919\n",
            "Epoch 127/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1362 - mae: 6.1362\n",
            "Epoch 128/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8467 - mae: 5.8467\n",
            "Epoch 129/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.1662 - mae: 6.1662\n",
            "Epoch 130/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8804 - mae: 5.8804\n",
            "Epoch 131/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9713 - mae: 5.9713\n",
            "Epoch 132/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8620 - mae: 5.8620\n",
            "Epoch 133/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8617 - mae: 5.8617\n",
            "Epoch 134/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.7643 - mae: 5.7643\n",
            "Epoch 135/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.7267 - mae: 5.7267\n",
            "Epoch 136/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.7747 - mae: 5.7747\n",
            "Epoch 137/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.7924 - mae: 5.7924\n",
            "Epoch 138/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0304 - mae: 6.0304\n",
            "Epoch 139/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.6596 - mae: 5.6596\n",
            "Epoch 140/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8270 - mae: 5.8270\n",
            "Epoch 141/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.7564 - mae: 5.7564\n",
            "Epoch 142/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.7339 - mae: 5.7339\n",
            "Epoch 143/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.7458 - mae: 5.7458\n",
            "Epoch 144/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.6799 - mae: 5.6799\n",
            "Epoch 145/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9922 - mae: 5.9922\n",
            "Epoch 146/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.7735 - mae: 5.7735\n",
            "Epoch 147/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.0076 - mae: 6.0076\n",
            "Epoch 148/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9545 - mae: 5.9545\n",
            "Epoch 149/150\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.9251 - mae: 5.9251\n",
            "Epoch 150/150\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8846 - mae: 5.8846\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Summarize the model"
      ],
      "metadata": {
        "id": "m51t2UfbxzI-"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "boston_regression_model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZaOiKPRdzigY",
        "outputId": "c15640db-290f-41d0-886b-afa2d8fcc929"
      },
      "execution_count": 52,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_13\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense_40 (Dense)            (None, 20)                280       \n",
            "                                                                 \n",
            " dense_41 (Dense)            (None, 10)                210       \n",
            "                                                                 \n",
            " dense_42 (Dense)            (None, 1)                 11        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 501 (1.96 KB)\n",
            "Trainable params: 501 (1.96 KB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Visualize the model training"
      ],
      "metadata": {
        "id": "e98wABSezm9B"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "pd.DataFrame(boston_regression_model_hist.history).plot()\n",
        "plt.ylabel(\"loss\")\n",
        "plt.xlabel(\"epochs\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 467
        },
        "id": "k3pNBLgtK7Rk",
        "outputId": "739da366-44de-4cbf-a47f-52d39a2989a6"
      },
      "execution_count": 84,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 0, 'epochs')"
            ]
          },
          "metadata": {},
          "execution_count": 84
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Evaluate the model"
      ],
      "metadata": {
        "id": "GmikgvwjzpuG"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred = boston_regression_model.predict(x_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "q3na6LECK7O-",
        "outputId": "d00ff7ce-d640-4287-ab38-dc8963e73b04"
      },
      "execution_count": 80,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4/4 [==============================] - 0s 4ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "boston_regression_model.evaluate(x_test, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6dRaLf1p10B9",
        "outputId": "e795ecdd-f153-43c4-e87c-a1efe90ada3f"
      },
      "execution_count": 85,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4/4 [==============================] - 0s 4ms/step - loss: 5.8236 - mae: 5.8236\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[5.823605537414551, 5.823605537414551]"
            ]
          },
          "metadata": {},
          "execution_count": 85
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.figure(figsize=(10, 7))\n",
        "plt.scatter(y_test, y_pred, c=\"b\", label=\"training data\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 617
        },
        "id": "H6KhxmQ31_aD",
        "outputId": "9fd76601-9681-4a38-8181-88a7efa7699f"
      },
      "execution_count": 86,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.collections.PathCollection at 0x7d6fe83d1870>"
            ]
          },
          "metadata": {},
          "execution_count": 86
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Model 2 - Without preprocessing\n",
        "\n",
        "1. Using Adam Optimizer"
      ],
      "metadata": {
        "id": "r8joD5kl3GPU"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "tf.random.set_seed(42)\n",
        "\n",
        "# Create a Model\n",
        "boston_regression_model_2 = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(15),\n",
        "    tf.keras.layers.Dense(10),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# Compile a Model\n",
        "boston_regression_model_2.compile(loss=tf.keras.losses.mae,\n",
        "                                optimizer=tf.keras.optimizers.Adam(learning_rate=0.01),\n",
        "                                metrics=[\"mae\"])\n",
        "\n",
        "# Fit a model\n",
        "boston_regression_model_hist_2 = boston_regression_model_2.fit(x_train, y_train, epochs=100)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1zLXTADk2DJl",
        "outputId": "623d0018-9f65-40c0-dc72-1cb37e7e316d"
      },
      "execution_count": 108,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "13/13 [==============================] - 1s 3ms/step - loss: 71.2896 - mae: 71.2896\n",
            "Epoch 2/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 22.6396 - mae: 22.6396\n",
            "Epoch 3/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 14.5127 - mae: 14.5127\n",
            "Epoch 4/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 8.5409 - mae: 8.5409\n",
            "Epoch 5/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.7341 - mae: 6.7341\n",
            "Epoch 6/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.1590 - mae: 6.1590\n",
            "Epoch 7/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.1734 - mae: 6.1734\n",
            "Epoch 8/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.4704 - mae: 6.4704\n",
            "Epoch 9/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.0883 - mae: 6.0883\n",
            "Epoch 10/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 7.9230 - mae: 7.9230\n",
            "Epoch 11/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.9761 - mae: 6.9761\n",
            "Epoch 12/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.9027 - mae: 5.9027\n",
            "Epoch 13/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.8886 - mae: 5.8886\n",
            "Epoch 14/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.8135 - mae: 5.8135\n",
            "Epoch 15/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.3858 - mae: 5.3858\n",
            "Epoch 16/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.8155 - mae: 6.8155\n",
            "Epoch 17/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 7.9351 - mae: 7.9351\n",
            "Epoch 18/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.7081 - mae: 6.7081\n",
            "Epoch 19/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.2309 - mae: 6.2309\n",
            "Epoch 20/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.9475 - mae: 5.9475\n",
            "Epoch 21/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.2541 - mae: 5.2541\n",
            "Epoch 22/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.2385 - mae: 5.2385\n",
            "Epoch 23/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.3848 - mae: 5.3848\n",
            "Epoch 24/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 8.1800 - mae: 8.1800\n",
            "Epoch 25/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.5112 - mae: 6.5112\n",
            "Epoch 26/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.3679 - mae: 5.3679\n",
            "Epoch 27/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 6.0544 - mae: 6.0544\n",
            "Epoch 28/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.6079 - mae: 5.6079\n",
            "Epoch 29/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.3064 - mae: 5.3064\n",
            "Epoch 30/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.4104 - mae: 5.4104\n",
            "Epoch 31/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.2834 - mae: 5.2834\n",
            "Epoch 32/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.4931 - mae: 6.4931\n",
            "Epoch 33/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.9652 - mae: 5.9652\n",
            "Epoch 34/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.5561 - mae: 6.5561\n",
            "Epoch 35/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.3262 - mae: 5.3262\n",
            "Epoch 36/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.7080 - mae: 4.7080\n",
            "Epoch 37/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.3280 - mae: 5.3280\n",
            "Epoch 38/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.3432 - mae: 5.3432\n",
            "Epoch 39/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.6220 - mae: 5.6220\n",
            "Epoch 40/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.7188 - mae: 5.7188\n",
            "Epoch 41/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.4054 - mae: 5.4054\n",
            "Epoch 42/100\n",
            "13/13 [==============================] - 0s 4ms/step - loss: 5.0005 - mae: 5.0005\n",
            "Epoch 43/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.1583 - mae: 5.1583\n",
            "Epoch 44/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.7171 - mae: 5.7171\n",
            "Epoch 45/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.6613 - mae: 5.6613\n",
            "Epoch 46/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.6976 - mae: 5.6976\n",
            "Epoch 47/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.6990 - mae: 5.6990\n",
            "Epoch 48/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.9915 - mae: 4.9915\n",
            "Epoch 49/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.7454 - mae: 4.7454\n",
            "Epoch 50/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.5993 - mae: 4.5993\n",
            "Epoch 51/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.5852 - mae: 4.5852\n",
            "Epoch 52/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.2427 - mae: 5.2427\n",
            "Epoch 53/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 7.3320 - mae: 7.3320\n",
            "Epoch 54/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.9202 - mae: 4.9202\n",
            "Epoch 55/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.8739 - mae: 4.8739\n",
            "Epoch 56/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.9037 - mae: 4.9037\n",
            "Epoch 57/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.1820 - mae: 5.1820\n",
            "Epoch 58/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.0856 - mae: 5.0856\n",
            "Epoch 59/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.9506 - mae: 4.9506\n",
            "Epoch 60/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.8900 - mae: 4.8900\n",
            "Epoch 61/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.5766 - mae: 4.5766\n",
            "Epoch 62/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.5803 - mae: 4.5803\n",
            "Epoch 63/100\n",
            "13/13 [==============================] - 0s 4ms/step - loss: 4.5749 - mae: 4.5749\n",
            "Epoch 64/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.7629 - mae: 4.7629\n",
            "Epoch 65/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.8359 - mae: 4.8359\n",
            "Epoch 66/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.9633 - mae: 4.9633\n",
            "Epoch 67/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.1813 - mae: 5.1813\n",
            "Epoch 68/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.2585 - mae: 5.2585\n",
            "Epoch 69/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.8539 - mae: 4.8539\n",
            "Epoch 70/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.5810 - mae: 4.5810\n",
            "Epoch 71/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.9452 - mae: 4.9452\n",
            "Epoch 72/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.4288 - mae: 4.4288\n",
            "Epoch 73/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.4660 - mae: 4.4660\n",
            "Epoch 74/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.5948 - mae: 4.5948\n",
            "Epoch 75/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.4799 - mae: 4.4799\n",
            "Epoch 76/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.4900 - mae: 4.4900\n",
            "Epoch 77/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.2950 - mae: 4.2950\n",
            "Epoch 78/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.3983 - mae: 5.3983\n",
            "Epoch 79/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 6.6644 - mae: 6.6644\n",
            "Epoch 80/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.3755 - mae: 4.3755\n",
            "Epoch 81/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.4902 - mae: 4.4902\n",
            "Epoch 82/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.4460 - mae: 4.4460\n",
            "Epoch 83/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.0324 - mae: 5.0324\n",
            "Epoch 84/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.6957 - mae: 4.6957\n",
            "Epoch 85/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.1849 - mae: 4.1849\n",
            "Epoch 86/100\n",
            "13/13 [==============================] - 0s 4ms/step - loss: 4.3075 - mae: 4.3075\n",
            "Epoch 87/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.1485 - mae: 4.1485\n",
            "Epoch 88/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.4205 - mae: 4.4205\n",
            "Epoch 89/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.9653 - mae: 4.9653\n",
            "Epoch 90/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.9626 - mae: 4.9626\n",
            "Epoch 91/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.0946 - mae: 4.0946\n",
            "Epoch 92/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.0944 - mae: 4.0944\n",
            "Epoch 93/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.7334 - mae: 4.7334\n",
            "Epoch 94/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.2576 - mae: 4.2576\n",
            "Epoch 95/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.3232 - mae: 4.3232\n",
            "Epoch 96/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.0289 - mae: 4.0289\n",
            "Epoch 97/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 3.9666 - mae: 3.9666\n",
            "Epoch 98/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.1816 - mae: 4.1816\n",
            "Epoch 99/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.2897 - mae: 4.2897\n",
            "Epoch 100/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.0523 - mae: 4.0523\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "boston_regression_model_2.evaluate(x_test, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "F261EUPq6FPv",
        "outputId": "b0df7386-89e5-4b15-c8d7-35e8446e931c"
      },
      "execution_count": 116,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4/4 [==============================] - 0s 4ms/step - loss: 4.2785 - mae: 4.2785\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[4.278454780578613, 4.278454780578613]"
            ]
          },
          "metadata": {},
          "execution_count": 116
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Learning Notes\n",
        "\n",
        "\n",
        "1.   Adam optimizer converges faster i.e. less number of epochs and less faster earning speed\n",
        "2.   The loss doesn't seem to improve much with tuning or changing of optimizers but efficiency in training changes\n",
        "3. Further going to try to train both models with normalized data"
      ],
      "metadata": {
        "id": "FX4nTqNY3lEb"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Normalize the data and preprocess"
      ],
      "metadata": {
        "id": "vZnCFuGW4TKH"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn import preprocessing\n",
        "\n",
        "X_train_normal = preprocessing.normalize(x_train)\n",
        "X_test_normal = preprocessing.normalize(x_test)"
      ],
      "metadata": {
        "id": "tRQlJtZT2FDG"
      },
      "execution_count": 100,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "X_train_normal.shape, x_train.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "g-IQd7N34xsg",
        "outputId": "f9152f84-6a60-43db-bf4a-be2fc667a3a4"
      },
      "execution_count": 105,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((404, 13), (404, 13))"
            ]
          },
          "metadata": {},
          "execution_count": 105
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Model 3 -\n",
        "1. Using model 2 as a baseline I will train the model again to see if loss reduces.\n"
      ],
      "metadata": {
        "id": "-QjQu0HN5frF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "tf.random.set_seed(42)\n",
        "\n",
        "\n",
        "# Create a Model\n",
        "boston_regression_model_3 = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(15),\n",
        "    tf.keras.layers.Dense(7),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# Compile a Model\n",
        "boston_regression_model_3.compile(loss=tf.keras.losses.mae,\n",
        "                                optimizer=tf.keras.optimizers.Adam(learning_rate=0.03),\n",
        "                                metrics=[\"mae\"])\n",
        "\n",
        "# Fit a model\n",
        "boston_regression_model_hist_3 = boston_regression_model_3.fit(X_train_normal, y_train, epochs=100)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BVdUiV5j4519",
        "outputId": "a69ed1a6-fc50-4c95-f7ad-f3216e44a8fb"
      },
      "execution_count": 135,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "13/13 [==============================] - 1s 4ms/step - loss: 14.4669 - mae: 14.4669\n",
            "Epoch 2/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 7.4301 - mae: 7.4301\n",
            "Epoch 3/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.6883 - mae: 5.6883\n",
            "Epoch 4/100\n",
            "13/13 [==============================] - 0s 4ms/step - loss: 5.3102 - mae: 5.3102\n",
            "Epoch 5/100\n",
            "13/13 [==============================] - 0s 4ms/step - loss: 5.1417 - mae: 5.1417\n",
            "Epoch 6/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 5.0110 - mae: 5.0110\n",
            "Epoch 7/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.0013 - mae: 5.0013\n",
            "Epoch 8/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.9598 - mae: 4.9598\n",
            "Epoch 9/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.1013 - mae: 5.1013\n",
            "Epoch 10/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.1079 - mae: 5.1079\n",
            "Epoch 11/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.0174 - mae: 5.0174\n",
            "Epoch 12/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 5.0347 - mae: 5.0347\n",
            "Epoch 13/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.7585 - mae: 4.7585\n",
            "Epoch 14/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.7114 - mae: 4.7114\n",
            "Epoch 15/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.6838 - mae: 4.6838\n",
            "Epoch 16/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.7525 - mae: 4.7525\n",
            "Epoch 17/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.7188 - mae: 4.7188\n",
            "Epoch 18/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.7815 - mae: 4.7815\n",
            "Epoch 19/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.5589 - mae: 4.5589\n",
            "Epoch 20/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.5124 - mae: 4.5124\n",
            "Epoch 21/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.5424 - mae: 4.5424\n",
            "Epoch 22/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.4789 - mae: 4.4789\n",
            "Epoch 23/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.3797 - mae: 4.3797\n",
            "Epoch 24/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.3707 - mae: 4.3707\n",
            "Epoch 25/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.7066 - mae: 4.7066\n",
            "Epoch 26/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.6116 - mae: 4.6116\n",
            "Epoch 27/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.4167 - mae: 4.4167\n",
            "Epoch 28/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.4201 - mae: 4.4201\n",
            "Epoch 29/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.3640 - mae: 4.3640\n",
            "Epoch 30/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.2774 - mae: 4.2774\n",
            "Epoch 31/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.2063 - mae: 4.2063\n",
            "Epoch 32/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.2117 - mae: 4.2117\n",
            "Epoch 33/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.1988 - mae: 4.1988\n",
            "Epoch 34/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.5677 - mae: 4.5677\n",
            "Epoch 35/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.3204 - mae: 4.3204\n",
            "Epoch 36/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.1232 - mae: 4.1232\n",
            "Epoch 37/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.3394 - mae: 4.3394\n",
            "Epoch 38/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.1108 - mae: 4.1108\n",
            "Epoch 39/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.2224 - mae: 4.2224\n",
            "Epoch 40/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.1052 - mae: 4.1052\n",
            "Epoch 41/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.2113 - mae: 4.2113\n",
            "Epoch 42/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.2614 - mae: 4.2614\n",
            "Epoch 43/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.9928 - mae: 3.9928\n",
            "Epoch 44/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.0786 - mae: 4.0786\n",
            "Epoch 45/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.0100 - mae: 4.0100\n",
            "Epoch 46/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.0417 - mae: 4.0417\n",
            "Epoch 47/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.0445 - mae: 4.0445\n",
            "Epoch 48/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.0248 - mae: 4.0248\n",
            "Epoch 49/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.9172 - mae: 3.9172\n",
            "Epoch 50/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.0240 - mae: 4.0240\n",
            "Epoch 51/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.9524 - mae: 3.9524\n",
            "Epoch 52/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.9775 - mae: 3.9775\n",
            "Epoch 53/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.3461 - mae: 4.3461\n",
            "Epoch 54/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.0146 - mae: 4.0146\n",
            "Epoch 55/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.9705 - mae: 3.9705\n",
            "Epoch 56/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.2097 - mae: 4.2097\n",
            "Epoch 57/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.2046 - mae: 4.2046\n",
            "Epoch 58/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.0887 - mae: 4.0887\n",
            "Epoch 59/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 4.3629 - mae: 4.3629\n",
            "Epoch 60/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.1346 - mae: 4.1346\n",
            "Epoch 61/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.9089 - mae: 3.9089\n",
            "Epoch 62/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.9523 - mae: 3.9523\n",
            "Epoch 63/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.8842 - mae: 3.8842\n",
            "Epoch 64/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.9017 - mae: 3.9017\n",
            "Epoch 65/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.1194 - mae: 4.1194\n",
            "Epoch 66/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.9702 - mae: 3.9702\n",
            "Epoch 67/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.0551 - mae: 4.0551\n",
            "Epoch 68/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.8815 - mae: 3.8815\n",
            "Epoch 69/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.8455 - mae: 3.8455\n",
            "Epoch 70/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.9381 - mae: 3.9381\n",
            "Epoch 71/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.7765 - mae: 3.7765\n",
            "Epoch 72/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.1291 - mae: 4.1291\n",
            "Epoch 73/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.1451 - mae: 4.1451\n",
            "Epoch 74/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.0128 - mae: 4.0128\n",
            "Epoch 75/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.8487 - mae: 3.8487\n",
            "Epoch 76/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.7790 - mae: 3.7790\n",
            "Epoch 77/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.7844 - mae: 3.7844\n",
            "Epoch 78/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 3.9013 - mae: 3.9013\n",
            "Epoch 79/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.1223 - mae: 4.1223\n",
            "Epoch 80/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.7256 - mae: 3.7256\n",
            "Epoch 81/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.7736 - mae: 3.7736\n",
            "Epoch 82/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.7617 - mae: 3.7617\n",
            "Epoch 83/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.8545 - mae: 3.8545\n",
            "Epoch 84/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.7647 - mae: 3.7647\n",
            "Epoch 85/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.9674 - mae: 3.9674\n",
            "Epoch 86/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.7450 - mae: 3.7450\n",
            "Epoch 87/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 3.7088 - mae: 3.7088\n",
            "Epoch 88/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.7647 - mae: 3.7647\n",
            "Epoch 89/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 4.0412 - mae: 4.0412\n",
            "Epoch 90/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 3.9083 - mae: 3.9083\n",
            "Epoch 91/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.8511 - mae: 3.8511\n",
            "Epoch 92/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.8147 - mae: 3.8147\n",
            "Epoch 93/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.7245 - mae: 3.7245\n",
            "Epoch 94/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.9253 - mae: 3.9253\n",
            "Epoch 95/100\n",
            "13/13 [==============================] - 0s 3ms/step - loss: 3.7790 - mae: 3.7790\n",
            "Epoch 96/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.6674 - mae: 3.6674\n",
            "Epoch 97/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.7466 - mae: 3.7466\n",
            "Epoch 98/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.7382 - mae: 3.7382\n",
            "Epoch 99/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.7129 - mae: 3.7129\n",
            "Epoch 100/100\n",
            "13/13 [==============================] - 0s 2ms/step - loss: 3.8775 - mae: 3.8775\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "boston_regression_model_3.evaluate(X_test_normal, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tUrvChNz6MBa",
        "outputId": "20a29ae0-c66e-45b4-d14d-2851e9e63719"
      },
      "execution_count": 136,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4/4 [==============================] - 0s 4ms/step - loss: 3.6682 - mae: 3.6682\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[3.668219566345215, 3.668219566345215]"
            ]
          },
          "metadata": {},
          "execution_count": 136
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred_3 = boston_regression_model.predict(X_test_normal)\n",
        "\n",
        "plt.figure(figsize=(10, 7))\n",
        "plt.scatter(y_test, y_pred_3, c=\"b\", label=\"training data\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 636
        },
        "id": "EI7oAgfE6qQi",
        "outputId": "5ef9b864-ace8-4cd2-f9d6-3126b8d94b0d"
      },
      "execution_count": 137,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4/4 [==============================] - 0s 3ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.collections.PathCollection at 0x7d6fe8a70520>"
            ]
          },
          "metadata": {},
          "execution_count": 137
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## TAKEAWAY ##\n",
        "\n",
        "1. Normalizing the data in boston housing price regression problem didn't helped that much to improve the accuracy of the model.\n",
        "2. Initial loss in models 1 & 2 was bigger as compared to model 3, that also proves normalized models learns and converges faster."
      ],
      "metadata": {
        "id": "Sj5hAQcJ7xQb"
      }
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "E43wT0Ib7twK"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}