tutorial109_regression_housing.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "tutorial109_regression_housing.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "mount_file_id": "1siy3I9R5O_P_lnwiWMXuAsRTIszUyV7h",
      "authorship_tag": "ABX9TyO1hqxZjqf1lcKp375N9DF7",
      "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/bnsreenu/python_for_image_processing_APEER/blob/master/tutorial109_regression_housing.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TTia-CEiCROV"
      },
      "source": [
        "https://youtu.be/vyX2_XRHO9g"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ioOS6r-vWZ2H"
      },
      "source": [
        "\"\"\"\n",
        "Boston housing dataset: https://www.cs.toronto.edu/~delve/data/boston/bostonDetail.html\n",
        "Data has white spaces, not commas.\n",
        "\n",
        "The Boston house-price data of Harrison, D. and Rubinfeld, D.L. 'Hedonic\n",
        " prices and the demand for clean air', J. Environ. Economics & Management,\n",
        " vol.5, 81-102, 1978\n",
        "\n",
        "    1. CRIM      per capita crime rate by town\n",
        "    2. ZN        proportion of residential land zoned for lots over \n",
        "                 25,000 sq.ft.\n",
        "    3. INDUS     proportion of non-retail business acres per town\n",
        "    4. CHAS      Charles River dummy variable (= 1 if tract bounds \n",
        "                 river; 0 otherwise)\n",
        "    5. NOX       nitric oxides concentration (parts per 10 million)\n",
        "    6. RM        average number of rooms per dwelling\n",
        "    7. AGE       proportion of owner-occupied units built prior to 1940\n",
        "    8. DIS       weighted distances to five Boston employment centres\n",
        "    9. RAD       index of accessibility to radial highways\n",
        "    10. TAX      full-value property-tax rate per $10,000\n",
        "    11. PTRATIO  pupil-teacher ratio by town\n",
        "    12. B        1000(Bk - 0.63)^2 where Bk is the proportion of blacks \n",
        "                 by town\n",
        "    13. LSTAT    % lower status of the population\n",
        "    14. MEDV     Median value of owner-occupied homes in $1000's\n",
        "\n",
        "\n",
        "\"\"\""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "prBUbeUqXTZ8"
      },
      "source": [
        "import pandas as pd\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense\n",
        "from keras.wrappers.scikit_learn import KerasRegressor\n",
        "from sklearn.model_selection import cross_val_score\n",
        "from sklearn.model_selection import KFold\n",
        "from sklearn.pipeline import Pipeline\n",
        "from sklearn.model_selection import train_test_split\n",
        "import seaborn as sns\n",
        "from matplotlib import pyplot as plt"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NLobcQ9GXVws",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "acd49341-67c7-4d71-e124-3001e5051d88"
      },
      "source": [
        "# load data and arrange into Pandas dataframe\n",
        "df = pd.read_csv(\"/content/drive/MyDrive/Colab Notebooks/data/housing_data/housing_data_for_regression.csv\", delim_whitespace=True, header=None)\n",
        "\n",
        "print(df.head())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "        0     1     2   3      4      5   ...  8      9     10      11    12    13\n",
            "0  0.00632  18.0  2.31   0  0.538  6.575  ...   1  296.0  15.3  396.90  4.98  24.0\n",
            "1  0.02731   0.0  7.07   0  0.469  6.421  ...   2  242.0  17.8  396.90  9.14  21.6\n",
            "2  0.02729   0.0  7.07   0  0.469  7.185  ...   2  242.0  17.8  392.83  4.03  34.7\n",
            "3  0.03237   0.0  2.18   0  0.458  6.998  ...   3  222.0  18.7  394.63  2.94  33.4\n",
            "4  0.06905   0.0  2.18   0  0.458  7.147  ...   3  222.0  18.7  396.90  5.33  36.2\n",
            "\n",
            "[5 rows x 14 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OnzyznIeXaVz",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6debfe9f-2fc4-49e1-c34a-85d757ba73e9"
      },
      "source": [
        "feature_names = ['CRIM', 'ZN', 'INDUS', 'CHAS', 'NOX', 'RM', 'AGE', \n",
        "                 'DIS', 'RAD', 'TAX', 'PTRATIO', 'B', 'LSTAT', 'PRICE']\n",
        "\n",
        "\n",
        "df.columns = feature_names\n",
        "print(df.head())\n",
        "\n",
        "#df = df.rename(columns={'MEDV': 'PRICE'})\n",
        "print(df.describe().T)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "      CRIM    ZN  INDUS  CHAS    NOX  ...    TAX  PTRATIO       B  LSTAT  PRICE\n",
            "0  0.00632  18.0   2.31     0  0.538  ...  296.0     15.3  396.90   4.98   24.0\n",
            "1  0.02731   0.0   7.07     0  0.469  ...  242.0     17.8  396.90   9.14   21.6\n",
            "2  0.02729   0.0   7.07     0  0.469  ...  242.0     17.8  392.83   4.03   34.7\n",
            "3  0.03237   0.0   2.18     0  0.458  ...  222.0     18.7  394.63   2.94   33.4\n",
            "4  0.06905   0.0   2.18     0  0.458  ...  222.0     18.7  396.90   5.33   36.2\n",
            "\n",
            "[5 rows x 14 columns]\n",
            "         count        mean         std  ...        50%         75%       max\n",
            "CRIM     506.0    3.613524    8.601545  ...    0.25651    3.677082   88.9762\n",
            "ZN       506.0   11.363636   23.322453  ...    0.00000   12.500000  100.0000\n",
            "INDUS    506.0   11.136779    6.860353  ...    9.69000   18.100000   27.7400\n",
            "CHAS     506.0    0.069170    0.253994  ...    0.00000    0.000000    1.0000\n",
            "NOX      506.0    0.554695    0.115878  ...    0.53800    0.624000    0.8710\n",
            "RM       506.0    6.284634    0.702617  ...    6.20850    6.623500    8.7800\n",
            "AGE      506.0   68.574901   28.148861  ...   77.50000   94.075000  100.0000\n",
            "DIS      506.0    3.795043    2.105710  ...    3.20745    5.188425   12.1265\n",
            "RAD      506.0    9.549407    8.707259  ...    5.00000   24.000000   24.0000\n",
            "TAX      506.0  408.237154  168.537116  ...  330.00000  666.000000  711.0000\n",
            "PTRATIO  506.0   18.455534    2.164946  ...   19.05000   20.200000   22.0000\n",
            "B        506.0  356.674032   91.294864  ...  391.44000  396.225000  396.9000\n",
            "LSTAT    506.0   12.653063    7.141062  ...   11.36000   16.955000   37.9700\n",
            "PRICE    506.0   22.532806    9.197104  ...   21.20000   25.000000   50.0000\n",
            "\n",
            "[14 rows x 8 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 334
        },
        "id": "zfyIBr17E7_x",
        "outputId": "e7e7b2d8-ead1-4456-ee4b-d27405952ed9"
      },
      "source": [
        "fig, ax =plt.subplots(1,3, figsize=(15, 5))\n",
        "sns.histplot(df['PRICE'], ax=ax[0])\n",
        "sns.histplot(df['AGE'], ax=ax[1])\n",
        "sns.histplot(df['TAX'], ax=ax[2])\n",
        "\n",
        "fig.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x360 with 3 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 334
        },
        "id": "lXpw5Ng-Fq45",
        "outputId": "004c9a2c-b7e0-4860-ba6f-f93a95d3e8e3"
      },
      "source": [
        "#Understand the data further. \n",
        "fig, ax =plt.subplots(1,3, figsize=(15, 5))\n",
        "sns.lineplot(x=df['AGE'], y=df['PRICE'], ax=ax[0])\n",
        "sns.lineplot(x=df['TAX'], y=df['PRICE'], ax=ax[1])\n",
        "sns.lineplot(x=df['B'], y=df['PRICE'], ax=ax[2])\n",
        "\n",
        "\n",
        "fig.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x360 with 3 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ExrbTajDGUaR",
        "outputId": "150b1848-77d7-4e3d-f7a1-b00f0dc43777"
      },
      "source": [
        "#Any strong correlation between attributes?\n",
        "corr = df.corr()\n",
        "corr_unstacked = corr.unstack()\n",
        "corr_unstacked_sorted = corr_unstacked.sort_values(kind=\"quicksort\", ascending=False)\n",
        "corr_df=pd.Series.to_frame(corr_unstacked_sorted, name='correlation')  #Convert Series to DataFrame and name correlation column accordingly. \n",
        "high_corr_features =  corr_df[corr_df.correlation != 1]  #Remove all 1s that correspond to self correlation\n",
        "print(high_corr_features.head(30))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "             correlation\n",
            "TAX   RAD       0.910228\n",
            "RAD   TAX       0.910228\n",
            "INDUS NOX       0.763651\n",
            "NOX   INDUS     0.763651\n",
            "      AGE       0.731470\n",
            "AGE   NOX       0.731470\n",
            "TAX   INDUS     0.720760\n",
            "INDUS TAX       0.720760\n",
            "PRICE RM        0.695360\n",
            "RM    PRICE     0.695360\n",
            "TAX   NOX       0.668023\n",
            "NOX   TAX       0.668023\n",
            "DIS   ZN        0.664408\n",
            "ZN    DIS       0.664408\n",
            "INDUS AGE       0.644779\n",
            "AGE   INDUS     0.644779\n",
            "CRIM  RAD       0.625505\n",
            "RAD   CRIM      0.625505\n",
            "NOX   RAD       0.611441\n",
            "RAD   NOX       0.611441\n",
            "INDUS LSTAT     0.603800\n",
            "LSTAT INDUS     0.603800\n",
            "      AGE       0.602339\n",
            "AGE   LSTAT     0.602339\n",
            "INDUS RAD       0.595129\n",
            "RAD   INDUS     0.595129\n",
            "NOX   LSTAT     0.590879\n",
            "LSTAT NOX       0.590879\n",
            "TAX   CRIM      0.582764\n",
            "CRIM  TAX       0.582764\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pjHOYk3uGsMx",
        "outputId": "d1218b45-e758-4c69-f08b-74760c8e87e1"
      },
      "source": [
        "print(df.isnull().sum())\n",
        "#df = df.dropna()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "CRIM       0\n",
            "ZN         0\n",
            "INDUS      0\n",
            "CHAS       0\n",
            "NOX        0\n",
            "RM         0\n",
            "AGE        0\n",
            "DIS        0\n",
            "RAD        0\n",
            "TAX        0\n",
            "PTRATIO    0\n",
            "B          0\n",
            "LSTAT      0\n",
            "PRICE      0\n",
            "dtype: int64\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TU6AE-61XbA3"
      },
      "source": [
        "#Split into features and target (Price)\n",
        "X = df.drop('PRICE', axis = 1)\n",
        "y = df['PRICE']\n",
        "\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 20)\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KOajsnJPXdWb"
      },
      "source": [
        "#Scale data, otherwise model will fail.\n",
        "#Standardize features by removing the mean and scaling to unit variance\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "scaler=StandardScaler()\n",
        "scaler.fit(X_train)\n",
        "\n",
        "X_train_scaled = scaler.transform(X_train)\n",
        "X_test_scaled = scaler.transform(X_test)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NAxSVZ6BXfXP",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "61e55474-05d1-4d19-ddc8-d83ac1a132ad"
      },
      "source": [
        "# define the model\n",
        "#Experiment with deeper and wider networks\n",
        "model = Sequential()\n",
        "model.add(Dense(128, input_dim=13, activation='relu'))\n",
        "model.add(Dense(64, activation='relu'))\n",
        "#Output layer\n",
        "model.add(Dense(1, activation='linear'))\n",
        "\n",
        "model.compile(loss='mean_squared_error', optimizer='adam', metrics=['mae'])\n",
        "model.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense (Dense)                (None, 128)               1792      \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 64)                8256      \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 1)                 65        \n",
            "=================================================================\n",
            "Total params: 10,113\n",
            "Trainable params: 10,113\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L2Pm9SkMXrbn",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6891dc09-5b81-4905-f191-71ad900b8a4f"
      },
      "source": [
        "history = model.fit(X_train_scaled, y_train, epochs =100, validation_data=(X_test_scaled, y_test))\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/100\n",
            "13/13 [==============================] - 1s 24ms/step - loss: 585.5664 - mae: 22.0591 - val_loss: 430.4849 - val_mae: 19.0240\n",
            "Epoch 2/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 447.1653 - mae: 18.9628 - val_loss: 330.0401 - val_mae: 16.3608\n",
            "Epoch 3/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 333.7310 - mae: 15.6798 - val_loss: 213.9000 - val_mae: 12.9314\n",
            "Epoch 4/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 195.6411 - mae: 11.8140 - val_loss: 109.5198 - val_mae: 8.7879\n",
            "Epoch 5/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 114.8703 - mae: 8.0923 - val_loss: 55.7692 - val_mae: 6.0671\n",
            "Epoch 6/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 59.2307 - mae: 5.8705 - val_loss: 38.1550 - val_mae: 4.9746\n",
            "Epoch 7/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 45.7845 - mae: 5.1300 - val_loss: 28.8134 - val_mae: 4.3029\n",
            "Epoch 8/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 32.1982 - mae: 4.0960 - val_loss: 24.6091 - val_mae: 3.8775\n",
            "Epoch 9/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 23.2372 - mae: 3.5271 - val_loss: 22.1117 - val_mae: 3.6475\n",
            "Epoch 10/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 27.9661 - mae: 3.6421 - val_loss: 20.9808 - val_mae: 3.5808\n",
            "Epoch 11/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 21.4870 - mae: 3.2684 - val_loss: 19.8817 - val_mae: 3.4529\n",
            "Epoch 12/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 24.2145 - mae: 3.3070 - val_loss: 18.8891 - val_mae: 3.3472\n",
            "Epoch 13/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 20.5714 - mae: 3.1170 - val_loss: 18.2481 - val_mae: 3.3044\n",
            "Epoch 14/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 17.4733 - mae: 2.9168 - val_loss: 17.5177 - val_mae: 3.2408\n",
            "Epoch 15/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 15.8686 - mae: 2.7973 - val_loss: 17.0261 - val_mae: 3.1948\n",
            "Epoch 16/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 16.5887 - mae: 2.9187 - val_loss: 16.5158 - val_mae: 3.1452\n",
            "Epoch 17/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 16.2183 - mae: 2.7637 - val_loss: 16.2595 - val_mae: 3.1042\n",
            "Epoch 18/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 16.2833 - mae: 2.7078 - val_loss: 15.8826 - val_mae: 3.0613\n",
            "Epoch 19/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 15.7747 - mae: 2.6728 - val_loss: 15.7366 - val_mae: 3.0667\n",
            "Epoch 20/100\n",
            "13/13 [==============================] - 0s 8ms/step - loss: 12.7315 - mae: 2.5050 - val_loss: 15.5801 - val_mae: 3.0369\n",
            "Epoch 21/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 13.9805 - mae: 2.5437 - val_loss: 15.3249 - val_mae: 3.0247\n",
            "Epoch 22/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 14.2894 - mae: 2.6940 - val_loss: 15.2164 - val_mae: 3.0028\n",
            "Epoch 23/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 12.8782 - mae: 2.4221 - val_loss: 15.0533 - val_mae: 2.9902\n",
            "Epoch 24/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 12.9013 - mae: 2.4677 - val_loss: 14.9247 - val_mae: 2.9727\n",
            "Epoch 25/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 13.1459 - mae: 2.4533 - val_loss: 14.8557 - val_mae: 2.9707\n",
            "Epoch 26/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 10.1635 - mae: 2.2588 - val_loss: 14.7736 - val_mae: 2.9547\n",
            "Epoch 27/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 11.2771 - mae: 2.3131 - val_loss: 14.5777 - val_mae: 2.9398\n",
            "Epoch 28/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 12.3352 - mae: 2.3847 - val_loss: 14.5927 - val_mae: 2.9283\n",
            "Epoch 29/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 8.7180 - mae: 2.0560 - val_loss: 14.3646 - val_mae: 2.9152\n",
            "Epoch 30/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 10.0950 - mae: 2.3395 - val_loss: 14.4156 - val_mae: 2.9108\n",
            "Epoch 31/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 9.7243 - mae: 2.1898 - val_loss: 14.3442 - val_mae: 2.8937\n",
            "Epoch 32/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 9.7885 - mae: 2.2558 - val_loss: 14.2081 - val_mae: 2.8961\n",
            "Epoch 33/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 10.4285 - mae: 2.2433 - val_loss: 14.1898 - val_mae: 2.8925\n",
            "Epoch 34/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 9.2665 - mae: 2.1658 - val_loss: 13.9170 - val_mae: 2.8785\n",
            "Epoch 35/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 12.9065 - mae: 2.3259 - val_loss: 13.8041 - val_mae: 2.8399\n",
            "Epoch 36/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 11.0385 - mae: 2.2743 - val_loss: 13.7923 - val_mae: 2.8533\n",
            "Epoch 37/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 10.1807 - mae: 2.1975 - val_loss: 13.8381 - val_mae: 2.8458\n",
            "Epoch 38/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 8.9221 - mae: 2.0748 - val_loss: 13.6556 - val_mae: 2.8334\n",
            "Epoch 39/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 9.5183 - mae: 2.1299 - val_loss: 13.5280 - val_mae: 2.8145\n",
            "Epoch 40/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 6.9544 - mae: 1.9405 - val_loss: 13.3796 - val_mae: 2.7993\n",
            "Epoch 41/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 8.7424 - mae: 2.1449 - val_loss: 13.6881 - val_mae: 2.7995\n",
            "Epoch 42/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 10.5392 - mae: 2.1662 - val_loss: 13.6158 - val_mae: 2.8073\n",
            "Epoch 43/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 10.0408 - mae: 2.1993 - val_loss: 13.6110 - val_mae: 2.8102\n",
            "Epoch 44/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 11.7907 - mae: 2.3209 - val_loss: 13.3536 - val_mae: 2.7835\n",
            "Epoch 45/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 9.5310 - mae: 2.1108 - val_loss: 13.4183 - val_mae: 2.7793\n",
            "Epoch 46/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 11.6218 - mae: 2.3014 - val_loss: 13.2006 - val_mae: 2.7568\n",
            "Epoch 47/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 7.3521 - mae: 2.0070 - val_loss: 13.2269 - val_mae: 2.7539\n",
            "Epoch 48/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 8.2680 - mae: 2.0801 - val_loss: 13.1878 - val_mae: 2.7456\n",
            "Epoch 49/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 10.5726 - mae: 2.2227 - val_loss: 13.1978 - val_mae: 2.7511\n",
            "Epoch 50/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 6.7919 - mae: 1.9185 - val_loss: 13.1680 - val_mae: 2.7400\n",
            "Epoch 51/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 10.0868 - mae: 2.1805 - val_loss: 13.3940 - val_mae: 2.7532\n",
            "Epoch 52/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 7.9048 - mae: 2.0217 - val_loss: 13.1064 - val_mae: 2.7199\n",
            "Epoch 53/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 7.4974 - mae: 1.9801 - val_loss: 13.3063 - val_mae: 2.7550\n",
            "Epoch 54/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 9.5415 - mae: 2.0444 - val_loss: 13.0602 - val_mae: 2.7202\n",
            "Epoch 55/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 8.2857 - mae: 1.9929 - val_loss: 13.0867 - val_mae: 2.7216\n",
            "Epoch 56/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 7.4801 - mae: 1.9506 - val_loss: 13.0603 - val_mae: 2.7134\n",
            "Epoch 57/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 7.5695 - mae: 2.0259 - val_loss: 13.0734 - val_mae: 2.7294\n",
            "Epoch 58/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 6.5609 - mae: 1.9043 - val_loss: 13.0008 - val_mae: 2.6915\n",
            "Epoch 59/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 8.1211 - mae: 2.0714 - val_loss: 13.0448 - val_mae: 2.7065\n",
            "Epoch 60/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 7.2293 - mae: 1.9242 - val_loss: 12.8728 - val_mae: 2.6858\n",
            "Epoch 61/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 6.3600 - mae: 1.8322 - val_loss: 13.0424 - val_mae: 2.6910\n",
            "Epoch 62/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 6.4127 - mae: 1.8563 - val_loss: 13.0509 - val_mae: 2.6979\n",
            "Epoch 63/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 9.0245 - mae: 2.1139 - val_loss: 12.8179 - val_mae: 2.6818\n",
            "Epoch 64/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 7.6520 - mae: 1.9697 - val_loss: 12.6286 - val_mae: 2.6597\n",
            "Epoch 65/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 5.6797 - mae: 1.7905 - val_loss: 13.0125 - val_mae: 2.7048\n",
            "Epoch 66/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 6.3483 - mae: 1.8757 - val_loss: 12.7335 - val_mae: 2.6604\n",
            "Epoch 67/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 6.4171 - mae: 1.8822 - val_loss: 12.8385 - val_mae: 2.6828\n",
            "Epoch 68/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 6.1415 - mae: 1.8534 - val_loss: 12.8276 - val_mae: 2.6564\n",
            "Epoch 69/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 6.4571 - mae: 1.8532 - val_loss: 12.9152 - val_mae: 2.6752\n",
            "Epoch 70/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 7.5514 - mae: 1.9013 - val_loss: 12.5596 - val_mae: 2.6398\n",
            "Epoch 71/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 5.8865 - mae: 1.7963 - val_loss: 12.6863 - val_mae: 2.6581\n",
            "Epoch 72/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 6.2926 - mae: 1.8201 - val_loss: 12.6707 - val_mae: 2.6453\n",
            "Epoch 73/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 6.3554 - mae: 1.8320 - val_loss: 12.5817 - val_mae: 2.6360\n",
            "Epoch 74/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 6.4989 - mae: 1.8666 - val_loss: 12.8120 - val_mae: 2.6702\n",
            "Epoch 75/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 7.8679 - mae: 1.9865 - val_loss: 12.7360 - val_mae: 2.6456\n",
            "Epoch 76/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 6.2021 - mae: 1.8359 - val_loss: 12.5809 - val_mae: 2.6424\n",
            "Epoch 77/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 5.6743 - mae: 1.7169 - val_loss: 12.4568 - val_mae: 2.6285\n",
            "Epoch 78/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 7.4854 - mae: 1.8464 - val_loss: 12.6377 - val_mae: 2.6421\n",
            "Epoch 79/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 6.0242 - mae: 1.7451 - val_loss: 12.7160 - val_mae: 2.6358\n",
            "Epoch 80/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 5.8903 - mae: 1.7389 - val_loss: 12.4848 - val_mae: 2.6138\n",
            "Epoch 81/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 6.6343 - mae: 1.8163 - val_loss: 12.6412 - val_mae: 2.6358\n",
            "Epoch 82/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 5.9693 - mae: 1.7613 - val_loss: 12.3982 - val_mae: 2.6012\n",
            "Epoch 83/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 5.9763 - mae: 1.7920 - val_loss: 12.3603 - val_mae: 2.6064\n",
            "Epoch 84/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 6.4132 - mae: 1.7685 - val_loss: 12.2535 - val_mae: 2.5956\n",
            "Epoch 85/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 5.1348 - mae: 1.7164 - val_loss: 12.3449 - val_mae: 2.5920\n",
            "Epoch 86/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 5.2271 - mae: 1.7151 - val_loss: 12.3193 - val_mae: 2.5928\n",
            "Epoch 87/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 4.9912 - mae: 1.6846 - val_loss: 12.6609 - val_mae: 2.6297\n",
            "Epoch 88/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 7.1042 - mae: 1.8174 - val_loss: 12.7775 - val_mae: 2.6600\n",
            "Epoch 89/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 5.3371 - mae: 1.7626 - val_loss: 12.5449 - val_mae: 2.6162\n",
            "Epoch 90/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 5.9113 - mae: 1.7006 - val_loss: 12.4678 - val_mae: 2.6080\n",
            "Epoch 91/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 5.8866 - mae: 1.7898 - val_loss: 12.5460 - val_mae: 2.6317\n",
            "Epoch 92/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 5.3984 - mae: 1.6885 - val_loss: 12.6435 - val_mae: 2.6203\n",
            "Epoch 93/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 5.5043 - mae: 1.7344 - val_loss: 12.6115 - val_mae: 2.6146\n",
            "Epoch 94/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 6.1751 - mae: 1.6916 - val_loss: 12.5680 - val_mae: 2.6219\n",
            "Epoch 95/100\n",
            "13/13 [==============================] - 0s 8ms/step - loss: 4.8762 - mae: 1.6811 - val_loss: 12.6638 - val_mae: 2.6236\n",
            "Epoch 96/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 5.0823 - mae: 1.7341 - val_loss: 12.1768 - val_mae: 2.5738\n",
            "Epoch 97/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 5.0828 - mae: 1.6175 - val_loss: 12.5445 - val_mae: 2.6211\n",
            "Epoch 98/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 4.6571 - mae: 1.5635 - val_loss: 12.5381 - val_mae: 2.6139\n",
            "Epoch 99/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 4.9745 - mae: 1.7105 - val_loss: 12.4966 - val_mae: 2.6036\n",
            "Epoch 100/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 4.6715 - mae: 1.5768 - val_loss: 12.6274 - val_mae: 2.6121\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tNuhndA-XtXj",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        },
        "outputId": "b7ba7daa-929e-41bb-fa18-ab3f645bb2f9"
      },
      "source": [
        "from matplotlib import pyplot as plt\n",
        "#plot the training and validation accuracy and loss at each epoch\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "epochs = range(1, len(loss) + 1)\n",
        "plt.plot(epochs, loss, 'y', label='Training loss')\n",
        "plt.plot(epochs, val_loss, 'r', label='Validation loss')\n",
        "plt.title('Training and validation loss')\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Loss')\n",
        "plt.legend()\n",
        "plt.show()\n",
        "\n",
        "\n",
        "acc = history.history['mae']\n",
        "val_acc = history.history['val_mae']\n",
        "plt.plot(epochs, acc, 'y', label='Training MAE')\n",
        "plt.plot(epochs, val_acc, 'r', label='Validation MAE')\n",
        "plt.title('Training and validation MAE')\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ecwEBzCvXvac",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "099ce795-84c1-46f6-b167-5919267574f8"
      },
      "source": [
        "#Predict on test data\n",
        "predictions = model.predict(X_test_scaled[:5])\n",
        "print(\"Predicted values are: \", predictions)\n",
        "print(\"Real values are: \", y_test[:5])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Predicted values are:  [[20.064781 ]\n",
            " [20.190065 ]\n",
            " [19.327238 ]\n",
            " [21.625889 ]\n",
            " [12.7529125]]\n",
            "Real values are:  498    21.2\n",
            "94     20.6\n",
            "150    21.5\n",
            "221    21.7\n",
            "423    13.4\n",
            "Name: PRICE, dtype: float64\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0ptG2e3SX2y2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "00e9f491-4797-41ec-bceb-37a8afc63498"
      },
      "source": [
        "#Comparison with other models..\n",
        "#Neural network - from the current code\n",
        "mse_neural, mae_neural = model.evaluate(X_test_scaled, y_test)\n",
        "print('Mean squared error from neural net: ', mse_neural)\n",
        "print('Mean absolute error from neural net: ', mae_neural)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "4/4 [==============================] - 0s 3ms/step - loss: 12.6274 - mae: 2.6121\n",
            "Mean squared error from neural net:  12.627412796020508\n",
            "Mean absolute error from neural net:  2.6121182441711426\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aDN4loIDX3Zc",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "df54e8a2-5336-421c-eaa2-6b5efca02872"
      },
      "source": [
        "#Linear regression\n",
        "from sklearn import linear_model\n",
        "from sklearn.tree import DecisionTreeRegressor\n",
        "from sklearn.metrics import mean_squared_error, mean_absolute_error\n",
        "\n",
        "### Linear regression\n",
        "lr_model = linear_model.LinearRegression()\n",
        "lr_model.fit(X_train_scaled, y_train)\n",
        "y_pred_lr = lr_model.predict(X_test_scaled)\n",
        "mse_lr = mean_squared_error(y_test, y_pred_lr)\n",
        "mae_lr = mean_absolute_error(y_test, y_pred_lr)\n",
        "print('Mean squared error from linear regression: ', mse_lr)\n",
        "print('Mean absolute error from linear regression: ', mae_lr)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mean squared error from linear regression:  16.49535197593167\n",
            "Mean absolute error from linear regression:  3.055894153890959\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J7Viu9PvX50z",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "53f9c624-99f9-47a4-bfbf-84213e8df095"
      },
      "source": [
        "#Random forest.\n",
        "#Increase number of tress and see the effect\n",
        "from sklearn.ensemble import RandomForestRegressor\n",
        "model = RandomForestRegressor(n_estimators = 30, random_state=30)\n",
        "model.fit(X_train_scaled, y_train)\n",
        "\n",
        "y_pred_RF = model.predict(X_test_scaled)\n",
        "\n",
        "mse_RF = mean_squared_error(y_test, y_pred_RF)\n",
        "mae_RF = mean_absolute_error(y_test, y_pred_RF)\n",
        "print('Mean squared error using Random Forest: ', mse_RF)\n",
        "print('Mean absolute error Using Random Forest: ', mae_RF)\n",
        "\n",
        "#Feature ranking...\n",
        "import pandas as pd\n",
        "feature_list = list(X.columns)\n",
        "feature_imp = pd.Series(model.feature_importances_, index=feature_list).sort_values(ascending=False)\n",
        "print(feature_imp)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mean squared error using Random Forest:  12.251427450980392\n",
            "Mean absolute error Using Random Forest:  2.382614379084967\n",
            "LSTAT      0.479179\n",
            "RM         0.337377\n",
            "DIS        0.065498\n",
            "CRIM       0.037599\n",
            "NOX        0.019502\n",
            "PTRATIO    0.014158\n",
            "TAX        0.012761\n",
            "B          0.012196\n",
            "AGE        0.010441\n",
            "INDUS      0.005591\n",
            "RAD        0.003762\n",
            "ZN         0.001591\n",
            "CHAS       0.000345\n",
            "dtype: float64\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}