ElasticNet_Regression.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyOKjX3B8Z2pO01DQCABwFyM",
      "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/sfansaria/Machine-Learning/blob/main/ElasticNet_Regression.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SRYO6I9eQlmk",
        "outputId": "c0ad0039-846f-4ac7-8817-3c42540824b8"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Import libraries\n",
        "\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split\n",
        "import matplotlib.pyplot as plt\n"
      ],
      "metadata": {
        "id": "sZaLwr4EQTRA"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#Elastic Net Regression\n",
        "\n",
        "class ElasticRegression():\n",
        "   def __init__(self,learning_rate, iterations, l1_penalty, l2_penalty):\n",
        "       self.learning_rate = learning_rate\n",
        "       self.iterations = iterations\n",
        "       self.l1_penalty = l1_penalty\n",
        "       self.l2_penalty = l2_penalty\n",
        "\n",
        "   #function for model learning\n",
        "   def fit(self, X, Y):\n",
        "       #no_of_training_examples, no_of_features\n",
        "\n",
        "       self.m, self.n = X.shape\n",
        "       #Weight initialization\n",
        "       self.W = np.zeros(self.n)\n",
        "       self.b = 0\n",
        "       self.X = X\n",
        "       self.Y = Y\n",
        "\n",
        "       #gradient descent learning\n",
        "       for i in range(self.iterations):\n",
        "           self.update_weights()\n",
        "       return self\n",
        "    #function to update weights in gradient descent\n",
        "\n",
        "   def update_weights(self):\n",
        "       Y_pred = self.predict(self.X)\n",
        "       #calculate gradients\n",
        "       dW = np.zeros(self.n)\n",
        "       for j in range(self.n):\n",
        "           if self.W[j] > 0:\n",
        "                dW[j] = (-(2*(self.X[:,j]).dot(self.Y-Y_pred)) + self.l1_penalty + 2*self.l2_penalty*self.W[j]) / self.m\n",
        "           else:\n",
        "                dW[j] = (-(2*(self.X[:,j]).dot(self.Y-Y_pred)) - self.l1_penalty + 2*self.l2_penalty*self.W[j]) / self.m\n",
        "\n",
        "\n",
        "       db = -2 * np.sum(self.Y - Y_pred) / self.m\n",
        "\n",
        "       #Update weights\n",
        "       self.W = self.W - self.learning_rate*dW\n",
        "       self.b = self.b - self.learning_rate*db\n",
        "\n",
        "       return self\n",
        "   #Hypothetical function h(x)\n",
        "\n",
        "   def predict(self, X):\n",
        "\n",
        "       return X.dot(self.W)+self.b\n",
        "\n",
        "def main():\n",
        "    #IMporting data set\n",
        "    df = pd.read_csv(\"/content/drive/My Drive/ml_datasets/salary_data.csv\")\n",
        "    X = df.iloc[:, :-1].values\n",
        "    Y = df.iloc[:, 1].values\n",
        "\n",
        "    #Splitting dataset into train and test set\n",
        "    X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.33,random_state=0)\n",
        "\n",
        "    #Model training\n",
        "    model = ElasticRegression(iterations=1000, learning_rate=0.01, l1_penalty=500, l2_penalty=1)\n",
        "    model.fit(X_train, Y_train)\n",
        "\n",
        "    #Prediction on test set\n",
        "    Y_pred = model.predict(X_test)\n",
        "    print(\"Predicted values\", np.round(Y_pred[:3], 2))\n",
        "    print(\"Real values     \", Y_test[:3])\n",
        "    print(\"Trained W       \", round(model.W[0], 2))\n",
        "    print(\"Trained b       \", round(model.b, 2))\n",
        "\n",
        "    #Visualization on test set\n",
        "    plt.scatter(X_test, Y_test, color=\"purple\")\n",
        "    plt.plot(X_test, Y_pred, color=\"blue\")\n",
        "    plt.title('Salary vs Experience')\n",
        "    plt.xlabel('Years of Experience')\n",
        "    plt.ylabel('Salary')\n",
        "    plt.show()\n",
        "\n",
        "\n",
        "if __name__ == \"__main__\":\n",
        "   main()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 541
        },
        "id": "BjAWYvrKQou1",
        "outputId": "056181e7-4cdd-42b2-f597-ee8a15401251"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Predicted values [ 40837.61 122887.43  65079.6 ]\n",
            "Real values      [ 37731. 122391.  57081.]\n",
            "Trained W        9323.84\n",
            "Trained b        26851.84\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    }
  ]
}