Regression_Multiple_Models_v1_1.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/IyadKhuder/Regression_Multiple_Models/blob/main/Regression_Multiple_Models_v1_1.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "##In this Notebook, I'm covering the following models: \n",
        "\n",
        "I.   Linear & Polynomial Regression \n",
        "\n",
        "II.   SVR\n",
        "\n",
        "III. Decision Tree\n",
        "\n",
        "IV. Random Forest Regression"
      ],
      "metadata": {
        "id": "s_CIwDo00g_d"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Case-study: \n",
        "An applicant for the job position \"ML engineer\" at IDC claimed that his previous gross salary per annum was $65k. Based on his CV, he is in the midway between junior and senior ML engineer.*\n",
        "\n",
        "Given the data sample of salary averages as a fenction of the position rankings, we're going to predict whether each of the two applicants is honest or bluffing.\n",
        " 🤔\n",
        "\n",
        "*Let's build a ML model that estimates the salary of that applicant, and predicts whether he was honest or just bluffing.*"
      ],
      "metadata": {
        "id": "nEPxXTnwMcjh"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n"
      ],
      "metadata": {
        "id": "3HjpZ_WC26rC"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EoRP98MpR-qj"
      },
      "source": [
        "## Importing the libraries"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd"
      ],
      "metadata": {
        "id": "EL6x9mr6dtGP"
      },
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fvBljkDj2pUN"
      },
      "source": [
        "## Connecting to your Google Drive"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4VY1J-m4fSeB",
        "outputId": "06cfc76f-9f1a-48e6-b652-3b08f104cd1d"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xpXdowrE9DxW"
      },
      "source": [
        "## Importing the dataset"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JjQJmIeVXI6X"
      },
      "source": [
        "### a) *Reading and storing the dataset in a dataframe, then verifying we access it*"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "Dataset_location_on_Google_Drive = \"/content/drive/MyDrive/ML AI DS/iyad_khuder_ML_training/Datasets/\"\n",
        "df = pd.read_csv(Dataset_location_on_Google_Drive+'Salaries_Position.csv')\n",
        "print(df)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mEECiqIKXDeQ",
        "outputId": "91a09beb-10c3-4dc5-ccfc-ce17b77bacff"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "                Position  Level  Gross_Salary_per_annum in USD\n",
            "0     Internship Trainee      1                          17000\n",
            "1    Junior Data Analyst      2                          38000\n",
            "2  Junior Data Scientist      3                          52000\n",
            "3         Senior Analyst      4                          58000\n",
            "4     Junior ML Engineer      5                          48000\n",
            "5     Senior ML Engineer      6                          65000\n",
            "6            Team Leader      7                          72000\n",
            "7        Project Manager      8                          92000\n",
            "8                    CEO      9                         158000\n",
            "9      Founder and owner     10                         299000\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3zlC9L2YXnKQ"
      },
      "source": [
        "### b) *Extracting the input and output fields*"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x= df.iloc[:,1:-1].values\n",
        "y= df.iloc[:,-1].values"
      ],
      "metadata": {
        "id": "CfaNVYZCMzKr"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MTgk_2pWmusP"
      },
      "source": [
        "### c) *Setting the printing options, so numbers are printed as whole numbers, withough scientific format or decimal digits.*"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "np.set_printoptions(suppress=True)\n",
        "np.set_printoptions(precision=6)\n",
        "np.set_printoptions(formatter={'float': '{: 0.0f}'.format})"
      ],
      "metadata": {
        "id": "EJ68wRkYeRL2"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rCZo6EAWnaJv"
      },
      "source": [
        "### d) *Preview*"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(x)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ixfjlAlJng_O",
        "outputId": "e415309e-00e9-49a6-c162-181216e29b1f"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[ 1]\n",
            " [ 2]\n",
            " [ 3]\n",
            " [ 4]\n",
            " [ 5]\n",
            " [ 6]\n",
            " [ 7]\n",
            " [ 8]\n",
            " [ 9]\n",
            " [10]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(y)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RhnlHJyRnnrN",
        "outputId": "19b3e3f1-2c87-4391-bb08-4709e4977209"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[ 17000  38000  52000  58000  48000  65000  72000  92000 158000 299000]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vN99YjPTDena"
      },
      "source": [
        "# I. Linear & Polynomial Regression"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4JyoqVLAe8dS"
      },
      "source": [
        "Training the Linear Regression model on the whole dataset"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.linear_model import LinearRegression\n",
        "lin_reg = LinearRegression()\n",
        "lin_reg.fit(x, y)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JCP-2p1KioGJ",
        "outputId": "acaf76f9-549d-47a9-a364-e4f0432eaa59"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LinearRegression()"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Rb5nWuSHEfBV"
      },
      "source": [
        "Training the Polynomial Regression model on the whole dataset"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "a) For deg = 3"
      ],
      "metadata": {
        "id": "JffkTHk3qGA9"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.preprocessing import PolynomialFeatures\n",
        "poly_reg_3 = PolynomialFeatures(degree=3)\n",
        "x_poly_3 = poly_reg_3.fit_transform(x)"
      ],
      "metadata": {
        "id": "Y6D0JJgAljbl"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(x_poly_3[:,2])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WFUuKJJAph08",
        "outputId": "0f8037ae-1505-457b-e71b-5b90bcc565e2"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[ 1  4  9  16  25  36  49  64  81  100]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x1_3 = x_poly_3[:,1]\n",
        "x2_3 = x_poly_3[:,2]\n",
        "x3_3 = x_poly_3[:,3]"
      ],
      "metadata": {
        "id": "TgItJv3tp8KH"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "The trick for the polynomila regression model is that it is converted into a multiple Linear Regression, where the powers of x are symbolized as multiple variables \"features\""
      ],
      "metadata": {
        "id": "4ChQtxb6s170"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "lin_reg_3 = LinearRegression()\n",
        "lin_reg_3.fit(x_poly_3, y)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vNy1G8Ecs171",
        "outputId": "d84eb3f2-fcb3-43ec-c8bd-6082e2f643e5"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LinearRegression()"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "b) For deg = 5"
      ],
      "metadata": {
        "id": "UcdWIz8wqM5G"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.preprocessing import PolynomialFeatures\n",
        "poly_reg_5 = PolynomialFeatures(degree=5)\n",
        "x_poly_5 = poly_reg_5.fit_transform(x)"
      ],
      "metadata": {
        "id": "18Q94z-OqM5H"
      },
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(x_poly_5[:,2])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b27809fb-871c-4349-8add-e515120ca831",
        "id": "LGBEu7MUqM5H"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[ 1  4  9  16  25  36  49  64  81  100]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x1_5 = x_poly_5[:,1]\n",
        "x2_5 = x_poly_5[:,2]\n",
        "x3_5 = x_poly_5[:,3]"
      ],
      "metadata": {
        "id": "nGpCCcJPqM5I"
      },
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "lin_reg_5 = LinearRegression()\n",
        "lin_reg_5.fit(x_poly_5, y)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1r5TcVt4sVDP",
        "outputId": "1f3a5e51-3e6b-4166-a162-46f6805e02c0"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LinearRegression()"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "c) For deg = 7"
      ],
      "metadata": {
        "id": "A_YhZC1P0kGL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.preprocessing import PolynomialFeatures\n",
        "poly_reg_7 = PolynomialFeatures(degree=7)\n",
        "x_poly_7 = poly_reg_7.fit_transform(x)"
      ],
      "metadata": {
        "id": "NCwGiYxL0kGY"
      },
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(x_poly_7[:,2])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0efa3659-8fd0-4cf2-e204-8fe4b92a8de3",
        "id": "K8nJAexq0kGY"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[ 1  4  9  16  25  36  49  64  81  100]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x1_7 = x_poly_7[:,1]\n",
        "x2_7 = x_poly_7[:,2]\n",
        "x3_7 = x_poly_7[:,3]"
      ],
      "metadata": {
        "id": "FoGcbsVb0kGZ"
      },
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "lin_reg_7 = LinearRegression()\n",
        "lin_reg_7.fit(x_poly_7, y)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "20702b8a-a3de-4612-b474-4541cdb1e2ab",
        "id": "DeYEc4be0kGZ"
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LinearRegression()"
            ]
          },
          "metadata": {},
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "stOnSo74E52m"
      },
      "source": [
        "## Visualising Linear & Polynomial Regression models all together for comparison"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Having tried degrees 3, 5 and 7, I'm going to just display the graphs for degrees 3 and 7. The graphs of the two degrees 5 and 7 aren't much different, so one of them suffices."
      ],
      "metadata": {
        "id": "3K_ivZtT3g02"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# plotting the real points of the dataset:\n",
        "plt.scatter(x, y, color = 'black')\n",
        "\n",
        "# plotting the linear model, as predicted by the trained LinearRegression:\n",
        "y_predict = lin_reg.predict(x)\n",
        "plt.plot(x, y_predict, color = 'grey', linestyle='dashed', label='Linear')\n",
        "\n",
        "# plotting the polynomial model of 3rd degree, \n",
        "y_predict_3 = lin_reg_3.predict(x_poly_3)\n",
        "plt.plot(x, y_predict_3, color = 'red', linestyle='dashed', label='Polynomial of 3rd deg.')\n",
        "\n",
        "# # plotting the polynomial model of 5th degree, \n",
        "# y_predict_5 = lin_reg_5.predict(x_poly_5)\n",
        "# plt.plot(x, y_predict_5, color = 'blue', label='Polynomial of 5th deg. (more accurate)')\n",
        "\n",
        "\n",
        "# plotting the polynomial model of 7th degree, \n",
        "y_predict_7 = lin_reg_7.predict(x_poly_7)\n",
        "plt.plot(x, y_predict_7, color = 'blue', label='Polynomial of 7th deg. (the most accurate)')\n",
        "\n",
        "plt.title('Truth or bluf? - Linear vs. Polynomial Regression')\n",
        "plt.xlabel('Job position level')\n",
        "plt.ylabel('Salary')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "w8VpjFvw52Tq",
        "outputId": "c6bd638a-9b62-4f4c-8127-89c8ea87f6a5"
      },
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "We notice the the polynomial model of 5th degree is the best fit of the graph. However, we should be mindful of the risk of falling into the trap pf an overfitting model."
      ],
      "metadata": {
        "id": "TNH8S59lt88M"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "diyJFZHhFFeK"
      },
      "source": [
        "## Predicting the salary based on the Linear Regression"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "np.set_printoptions(precision=2)"
      ],
      "metadata": {
        "id": "LwP_htXUsUIy"
      },
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "x0 = [[5.5]]"
      ],
      "metadata": {
        "id": "7Sjz1LhkrLgY"
      },
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "y0_lin = lin_reg.predict(x0)"
      ],
      "metadata": {
        "id": "Zi8CzMqjr8nn"
      },
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(y0_lin)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hgyvhfH--xAf",
        "outputId": "4c3efd59-4185-45f9-ec13-2d44fc34fc86"
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[89900.]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Outcome:** \n",
        "\n",
        "According to the linear regression model, a salary of $65k is very modest for such an engineer and he deserves much more than that! 😇"
      ],
      "metadata": {
        "id": "ONgf5UjwI77Y"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DW7I7ZVDFNkk"
      },
      "source": [
        "## Predicting the salary based on the Polynomial Regression"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x0_pol = poly_reg_7.fit_transform(x0)"
      ],
      "metadata": {
        "id": "RoPaD5FGslYE"
      },
      "execution_count": 26,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "y7_pol = lin_reg_7.predict(x0_pol)"
      ],
      "metadata": {
        "id": "R4pj6BQis4zz"
      },
      "execution_count": 27,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(y7_pol)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ScRE4_mZ-6Az",
        "outputId": "2d644ffc-810d-405d-c519-f9c5f2ce1975"
      },
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[57004.88]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Outcome:** \n",
        "\n",
        "According to the polynomial regression model of 7th degree (which is much more accurate than the linear one), the estmated salary for this applicant is approx. $57 thousands, which is pretty lower than the number claimed by the applicant, and this means that he's likely to be bluffing 🤥"
      ],
      "metadata": {
        "id": "Fy_SNEMXh8Xh"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## **Conclusion:** \n",
        "\n",
        "For non-monotone data, which has severe fluctuations, that is, in mathematical words, outstanding local minimum/maximum points, the linear regression fails and it's needed to use a more sophisticated model, such as the polynomial one. \n",
        "\n",
        "It's noticed that the polynomial model of 7th degree beats its counterpart of 3rd degree at several points, such as at x=6.\n",
        "\n",
        "However, we should carefully select the optimal degree. Sometimes, going a too high degree may result in overfitting, which will have a negative impact on the prediction of new values.\n"
      ],
      "metadata": {
        "id": "hYD5TTl5JjnI"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n",
        "---\n",
        "\n"
      ],
      "metadata": {
        "id": "__4ER9pjwhf4"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "m3PAEPRDRLA3"
      },
      "source": [
        "# II. Support Vector Regression (SVR)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "piJwIv7NwwWv"
      },
      "source": [
        "*Data Preprocessing*"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YS8FeLHYS-nI"
      },
      "source": [
        "Feature Scaling"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JfCTSX6RyDQJ",
        "outputId": "d1c8de88-7783-42cc-d896-a151a0a6c0f9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(10, 1)"
            ]
          },
          "metadata": {},
          "execution_count": 150
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y=y.reshape(len(y),1)\n",
        "y.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3ZKc77wayFXh",
        "outputId": "66b73213-50fa-4db9-e978-62c4dc429b82"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(10, 1)"
            ]
          },
          "metadata": {},
          "execution_count": 151
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Note: without the above command\n",
        ", we'd get the error\n",
        "<br> \"Expected 2D array, got 1D array instead\"\n",
        " <br>  for the y array y, when we try the command \n",
        "<br> y_scaled = sc_y.fit_transform(y)\n",
        "<br> in the cell below\n",
        " "
      ],
      "metadata": {
        "id": "vhxWVLsjxsaj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.preprocessing import StandardScaler\n",
        "sc_x = StandardScaler()\n",
        "x_scaled = sc_x.fit_transform(x)\n",
        "sc_y = StandardScaler()\n",
        "y_scaled = sc_y.fit_transform(y)"
      ],
      "metadata": {
        "id": "sg7U3LKhV6Lo"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(x_scaled)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "y70BAaXgXEUn",
        "outputId": "72679967-e3af-4791-865a-a10d0f86c4e6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[-2]\n",
            " [-1]\n",
            " [-1]\n",
            " [-1]\n",
            " [-0]\n",
            " [ 0]\n",
            " [ 1]\n",
            " [ 1]\n",
            " [ 1]\n",
            " [ 2]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(y_scaled)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "03i-R317XLk0",
        "outputId": "41a0fb29-6601-451c-8986-6c4df53ed4f6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[-1]\n",
            " [-1]\n",
            " [-1]\n",
            " [-0]\n",
            " [-0]\n",
            " [-0]\n",
            " [-0]\n",
            " [ 0]\n",
            " [ 1]\n",
            " [ 3]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eiU6D2QFRjxY"
      },
      "source": [
        "Training the SVR model on the whole dataset"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.svm import SVR\n",
        "regressor = SVR(kernel = 'rbf')\n",
        "regressor.fit(x_scaled,y_scaled)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZXLeQzVjXrtG",
        "outputId": "397f9301-2368-4225-dc98-19af837f36a4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:993: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
            "  y = column_or_1d(y, warn=True)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SVR()"
            ]
          },
          "metadata": {},
          "execution_count": 155
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "deDnDr8UR5vq"
      },
      "source": [
        "*Predicting the output for a single value*"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "The steps to follow here are:\n",
        "\n",
        "\n",
        "1.   Scale the input value (position level = 6.5), using the same object, which was used for x; i.e. sc_x\n",
        "2.   Apply the prediction method on the scaled value\n",
        "2.   The obtained value of y will be in the scaled form. Therefore, we need to apply the inverse transformation on it to get the corresponding value.\n",
        "\n",
        "Important note:\n",
        "For the requested value 6.5,\n",
        "when we want to scale it, we should use the function sc_x.transform to scale it, NOT sc_x.fit_transform\n",
        "if we use the latter one, it will override the original scaling for the array x, and scale it according to 6.5\n",
        "and in that case, the scaled value will be 0, because:\n",
        "for a single value, the mean is the same as the value itself, and therefore, the scaled value will be (6.5-6.5)/sigma = 0\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "64kvxLiTHWZN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "x0= np.array([[5.5]])"
      ],
      "metadata": {
        "id": "M6LKzCHmTmC0"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "x0_scaled = sc_x.transform(x0)"
      ],
      "metadata": {
        "id": "3gSxu6mqWdM_"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(x0_scaled)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PuiFCTv4W5wb",
        "outputId": "f40b4052-dd50-4828-e369-1628de45ef59"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[ 0]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y0_scaled = regressor.predict(x0_scaled)\n",
        "print(y0_scaled)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "H8Gw-Bz8J-us",
        "outputId": "3c6712ba-79dc-4b20-e8d2-872f61983d16"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[-0]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y0 = sc_y.inverse_transform([y0_scaled])\n",
        "print(y0)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rKZYQpxKJHVZ",
        "outputId": "57b72595-0bb0-42a0-b5e2-55118a19bc24"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[ 57473]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Note on the scaling:\n",
        "When we scale a sample, all the scaled values will be in the range [approx. -2.7, approx. 2.7],\n",
        "Having applied fit_transform, if we scale a new value, which is beyond the range of the given sample, the scaled value will also be beyond the standard range mentioned above.\n",
        "In ML models, we usually try to find such extreme values.\n",
        "For example, the middle value in the array below, will be scaled to a value that is greater than 10000."
      ],
      "metadata": {
        "id": "yWyi5t4yBOe1"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "x2= np.array([[2,  60000, 7.6]])\n",
        "x2col = np.transpose(x2)\n",
        "# x0col = np.transpose(x0)\n",
        "x2_scaled = sc_x.transform(x2col)\n",
        "print(x2_scaled)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "C_gMi0AyYoMu",
        "outputId": "fc737e6b-83a1-458a-bc10-1d85bacb60f3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[-1]\n",
            " [ 20887]\n",
            " [ 1]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zzedFlUISSu_"
      },
      "source": [
        "*Visualising the SVR results*"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Getting the predicted values for the y, based on the trained model:\n",
        "y_predict_scaled = regressor.predict(x_scaled)\n",
        "print(y_predict_scaled)\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ccdf2a93-0bf1-4d1b-b6d5-54681d015de0",
        "id": "SWqArDgYxBmU"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[-1 -1 -1 -0 -0 -0 -0  0  1  1]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_predict_scaled.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eiJ0gwa4MSne",
        "outputId": "b79612b3-f8d8-4441-dec5-8b69664bb63c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(10,)"
            ]
          },
          "metadata": {},
          "execution_count": 164
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_predict = sc_y.inverse_transform([y_predict_scaled])\n",
        "y_predict.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fsiI0HphMluZ",
        "outputId": "ac7f90e3-59d3-4efa-d993-6d498486435e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1, 10)"
            ]
          },
          "metadata": {},
          "execution_count": 165
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_predict_v = np.transpose(y_predict)"
      ],
      "metadata": {
        "id": "pDObtq5PN5oU"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "y_predict_v.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oIsdMsZTM2RN",
        "outputId": "5319269a-d593-40ee-b2c3-b49ad058a81a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(10, 1)"
            ]
          },
          "metadata": {},
          "execution_count": 167
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# plotting the real points of the dataset:\n",
        "plt.scatter(x, y, color = 'black')\n",
        "# plotting the SVR model, as predicted by the trained SVR Regression:\n",
        "\n",
        "\n",
        "\n",
        "plt.plot(x, y_predict_v, color = 'blue', linestyle='dashed', label='SVR Regression')\n",
        "\n",
        "plt.title('Truth or bluf? ')\n",
        "plt.xlabel('Job position level')\n",
        "plt.ylabel('Salary')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "SUAy74iuHOcc",
        "outputId": "651c0b7e-d86f-4783-b927-4a1df2e1d457"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n",
        "---\n",
        "\n"
      ],
      "metadata": {
        "id": "kz0SkGK63kZQ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# III. Decision Tree"
      ],
      "metadata": {
        "id": "NaNnMYA63tJT"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "g16qFkFQVC35"
      },
      "source": [
        "*Training the Decision Tree Regression model on the whole dataset*"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SLDKyv1SVUqS",
        "outputId": "8286df32-2b22-4cba-8f98-1740c62431a8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "from sklearn.tree import DecisionTreeRegressor\n",
        "# In the following line, we precise the seed of the random function, so we get the same pattern everytime we run the code\n",
        "TreeRegressor = DecisionTreeRegressor(random_state = 0)\n",
        "TreeRegressor.fit(x, y)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "DecisionTreeRegressor(random_state=0)"
            ]
          },
          "metadata": {},
          "execution_count": 169
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_3FRdl6bje9N"
      },
      "source": [
        "*Predicting the output for a single value*"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_FpGZf7vVgrK",
        "outputId": "8ba7f210-afd8-4558-f436-fb3e700ca8e6",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "TreeRegressor.predict([[5.5]])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ 48000])"
            ]
          },
          "metadata": {},
          "execution_count": 170
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ph8ExBj0VkIT"
      },
      "source": [
        "*Visualising the Decision Tree Regression results (higher resolution)*"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zzH1Vv1oVrqe",
        "outputId": "8f3966ec-19c1-497e-9b42-ac80667d27bc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "x_grid = np.arange(min(x), max(x), 0.1)\n",
        "x_grid = x_grid.reshape((len(x_grid), 1))\n",
        "plt.scatter(x, y, color = 'red')\n",
        "plt.plot(x_grid, TreeRegressor.predict(x_grid), color = 'blue')\n",
        "plt.title('Truth or Bluff (Decision Tree Regression)')\n",
        "plt.xlabel('Position level')\n",
        "plt.ylabel('Salary')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZcAAAEWCAYAAACqitpwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3de5zVVb3/8ddbERVFkSRDFMeCLtjPTElNO+VdNAvr50mLEjsmXbSy00XNfkfNKLOTlsdL4SXRSCWzpLSjqNhF08BLKlpBpgiiooOCDajg5/fHWlv2DLNn9h6+w9575v18PPZj71nfy1rfPTP7s9flu5YiAjMzsyJtUO8CmJlZ3+PgYmZmhXNwMTOzwjm4mJlZ4RxczMyscA4uZmZWOAcX65akxyQdUO9ylEgKSaOq3HdvSfMkvSjpcEnbSPq9pOWSvl/hmIMl/arYUlcs31xJ+3Szz8hc/g3XR5msMkkTJN1cwHm2kfSIpI2LKFcjcnDpA/IHT+nxqqQVZT9PqPFcl0v6Vm+VtYr898nXUCr/IklnrMMpvwmcHxGbR8SvgEnAs8AWEfHlCsdMBs4qK1NI+lcuz3OSbpV05DqU6TURsVNE3N7NPgty+VcXkSe0C1ilR/k1vijp34rKq5O8y/NaJOmcZgmcETEtIg4q4DxPA7NIf499koNLH5A/eDaPiM2BBcAHytKmlfaTNKB+pVxbF+V5sux63gMcK+nwHmazAzC3w88PR4W7hyW9C9gyIu7qsOkduTxvAS4Hzpd0Wg/LVHdlAav0PkO+xvz4Q2nfXvq7Kb2f7wOOBP6j6Awa7e+9E9OAT9e7EL3FwaUPy7WAhZJOkvQU8BNJx0j6Y4f9QtIoSZOACcDX8rfKX5fttoukByS9IOkaSZtUyHMDSd+Q9LikZyRdIWnLvK0l53WspAXAbd1dQ0T8E7gTGFMhv9slfars59euT9I/gDcCv87XcxUwsez6OmvqOwT4XRfleTYirgQ+C5wi6XU5ry0lXSppcf42/q3yb+OSjsvNIMslPSxp15z+WpOjpN0lzZG0TNLTks7p8L4NyD9vK2mGpFZJ8yUdV5bP6ZKm5/d9eW52G9vd+9zhPT1G0h2SzpX0HHC6pI0l/bekBblsP5K0adkxh0m6X9Lzku6UtHM1eUXEfOAOYJdqziVpV0n35Wv7ef5b/Fbe1tnf+waSTpb0j1zrnC5paN5/E0k/zenPS5otaZuy9+DRnM8/lVsAOv7/SNorH/dCft6rbNvtks7M7+VySTdL2rrs8u8G3ihph1p+P83CwaXvewMwlPSNvcsqeERMIX2bOjt/e/1A2eaPAOOAHYGdgWMqnOaY/NiX9MG+OXB+h33eB7wNOLi7wksaDewNdKxJdCsi3kT7mtxHaX99t3Ry2P8B/lbF6a8HBgC7558vB1YBo4B3AgcBn8rX8O/A6cDRwBbAB4HnOjnnD4EfRsQWwJuA6RXyvhpYCGwLHAF8W9J+Zds/mPcZAsxg7fe/GnsAjwLbsKaZ8M2kIDAKGAH8V76+dwKXkb6Fvw74MTBDVfQnSHor8G/A/O7OJWkg8EvSez0UuAr4UIdTdvx7/zxwOOlvbltgKXBB3ncisCWwfc7rM8AKSZsB5wGHRMRgYC/g/k7KPhS4Ie/7OuAc4IbSF47sY8AngdcDA4GvlDZExKp83e/o7n1qRg4ufd+rwGkR8VJErFiH85wXEU9GRCvwa8q+aXYwATgnIh6NiBeBU4Cj1L6J4vSI+FcX5dk2f5NcBvyd9A3vjxX2LdoQYHl3O0XEK6S+m6H52+6hwIn5up4BzgWOyrt/ihTQZkcyPyIe7+S0rwCjJG0dES920jSHpO1JwfakiFgZEfcDl5ACV8kfI+LG3EdzJT378HoyIv4nfwCuJH1QfykiWiNiOfDtsuubBPw4Iu6OiNURMRV4Cdizi/PfK+lfwCPA7cCFVZxrT1JAPy8iXomI64A/dzhvx7/3zwCnRsTCiHiJFOSPyH+Pr5CCwqic1z0RsazsPG+XtGlELI6Iuazt/cC8iLgyIlZFxFXAX4HyL2U/iYi/57JMZ+3/m+Wkv7k+x8Gl71sSESsLOM9TZa/bSDWSzmwLlH9wPk76QNimLO2JbvJ6MiKG5G/wQ4AVwNTaittjS4HB3e0kaSNgGNBK+pa8EbA4B8XnSd+4X5933x74RxV5H0uqHfw1N7Ec1sk+2wKlD/iSx0k1iZKOv6tNVHv/Q/nvaBgwCLin7Pr+N6dDuv4vl7bl7dvnslayK+lv6EhSLWmzKs61LbCoQ39Zx7+ljn/vOwC/LDvXI8Bq0t/jlcBNwNWSnpR0tqSNIuJfuVyfIf1Ob8g1rI46/q1D97+Ljv83g4HnOzl303Nw6fs6dlz/i/RBAYCkN3Szf62eJP1Dl4wkNRc93ZM8IuIF4Ge0/zZYrt31kJpF1sUDpA/47ownXdefSR9wLwFb56A4JCK2iIid8r5PkJq5uhQR83LT3euB7wLX5iaack+SakvlAXAksKiKMtei/Hf0LCnA71R2fVuWDQR4Aphctm1IRAzK3+QrZ5BMB/5EbmLr5lyLgRGSVHaa7bsod+l8h3Q43yYRsSjXfs6IiDGkpq/DyDXAiLgpIg4EhpNqIxd3cgkd/9ahht9FDvijgL9Us3+zcXDpf/4C7CRpF6VO+dM7bH+a1FfSU1cBX5K0o6TNSc0n1+TmlZrlcxxF+xFf5e4HPixpkNK9L8f2JJ8yN5La5yuVZ2ju3L0A+G5EPBcRi4Gbge9L2iJ3Ir9JUuk8lwBfkbSbklGddeJK+rikYRHxKmu+zb5avk9EPEEa4PCd3CG9c77mn67jdVeUy3MxcK6k1+eyjpBU6jO7GPiMpD3y9W0m6f0dAmBXzgKOy190ujrXn0i1jhMkDZA0njV9XpX8CJhcer8lDcvHIWlfSf9HaeDFMlIz2atK96CMz4H9JeBFOvweshuBN0v6WC7PkaSBJ7+p8rp3Bx6r0ETa9Bxc+pmI+Dvp3o9bgHms3ZdxKTAmNyP05EbCy0jNDb8H/klqr/98jefYVvl+C1Izw1BSX05nzgVeJgXFqaQO+x6LiHuBFyTt0WHTX3J55pP6UL4UEf9Vtv1oUoftw6SmtWtJ33qJiJ+TOsV/Rmpj/1W+po7GAXNzPj8EjqrQL/VRoIX0zfmXpD6GzgYnFOkk0rXflfvCbiENyyYi5gDHkQYOLM37HVPtiSPiQdLfy1e7OldEvAx8mBRMnwc+Tvogf6mL0/+QNKjhZknLSQNDSr/bN5B+T8tIzWW/I/3tbgD8J+n9bSV92fhsJ+V+jlTb+TJpgMbXgMMi4tkqL30CKfj1SQovFmbWjqSDgM9FRE/vrbH1RNLdwI8i4if1Lkstcg3wd8A7C+oTbTgOLmbWNHJT499I/UClb/5vzE2T1kAa/Q5WM7NybyEN6d2MdB/OEQ4sjck1FzMzK5w79M3MrHBuFsu23nrraGlpqXcxzMyayj333PNsRAzrmO7gkrW0tDBnzpx6F8PMrKlI6vQ+HTeLmZlZ4RxczMyscA4uZmZWOAcXMzMrXK8Flzyp3p8l/UVpNbwzcvqOku5WWkHvmrwAEHkxoGty+t2SWsrOdUpO/1vZZHlIGpfT5ks6uSy90zzMzGz96M2ay0vAfhHxDtICOeMk7UmaSvzciBhFmpiuNIvtscDSnH5u3g9JY0iz4u5EmtjvQkkb5plMLyAtSzsG+Gjely7yMDOzkmnToKUFNtggPU9bp3lf2+m14JLXangx/7hRfgSwH2kmUkiz2JYmBxzPmgWhrgX2z+s2jAeuzivL/ZM0S+ru+TE/0oqHL5OWdR2fj6mUh5mZQQokkybB449DRHqeNKmwANOr97nk2sU9pAVxLiCtxvd82doeC1mzatsI8qpyEbFK0gukJUhH0H799PJjnuiQvkc+plIeZmZ90h13wE03Vbfv4MHw1QtOhba29hva2uDUU2FCpRUuqterwSWv4b2LpCGkdSc6Wyq0biRNIq3ZzciRI+tcGjOznvvGN+D226HdOp0VDB8OX128oPONCyqk12i9jBaLiOeBWcC7gSFas573dqxZEnQRecnSvH1L0gI8r6V3OKZS+nNd5NGxXFMiYmxEjB02bK3ZC8zMmsaLL8Ihh8Crr3b/WLQIqPSFuqAv2r05WmxYrrEgaVPgQNJqb7OAI/JuE4Hr8+sZ+Wfy9tsiTdk8AzgqjybbERhNWrd8NjA6jwwbSOr0n5GPqZSHmVmftGIFbLJJDQdMngyDBrVPGzQopRegN5vFhgNTc7/LBsD0iPiNpIeBqyV9C7iPtKwu+flKSfNJS4seBRARcyVNJy0fuwo4Pje3IekE4CZgQ+CyiCits35ShTzMzPqkFStg001rOKDUr3LqqakpbOTIFFgK6G8Br+fymrFjx4YnrjSzZjViRGoWu+SS9ZuvpHsiYmzHdN+hb2bWB9Rcc+llDi5mZn1AzX0uvczBxcysyUXAypWuuZiZWYFeeik9O7iYmVlhVqxIzw4uZmZWmFJwcZ+LmZkVxjUXMzMr3MqV6dnBxczMCuOai5mZFc7BxczMCucOfTMzK5z7XMzMrHBuFjMzs8I5uJiZWeHc52JmZoVzn4uZmRXOzWJmZla4FStAgoED612SNRxczMyaXGmhMKneJVnDwcXMrMk12kJh4OBiZtb0VqxwcDEzs4I5uJiZWeFKfS6NxMHFzKzJuc/FzMwK52YxMzMrXL8KLpK2lzRL0sOS5kr6Yk4/XdIiSffnx6Flx5wiab6kv0k6uCx9XE6bL+nksvQdJd2d06+RNDCnb5x/np+3t/TWdZqZ1Vu/Ci7AKuDLETEG2BM4XtKYvO3ciNglP24EyNuOAnYCxgEXStpQ0obABcAhwBjgo2Xn+W4+1yhgKXBsTj8WWJrTz837mZn1Sf2qQz8iFkfEvfn1cuARYEQXh4wHro6IlyLin8B8YPf8mB8Rj0bEy8DVwHhJAvYDrs3HTwUOLzvX1Pz6WmD/vL+ZWZ/Tbzv0c7PUO4G7c9IJkh6QdJmkrXLaCOCJssMW5rRK6a8Dno+IVR3S250rb38h729m1uf0t2YxACRtDvwCODEilgEXAW8CdgEWA9/v7TJ0UbZJkuZImrNkyZJ6FcPMbJ30u+AiaSNSYJkWEdcBRMTTEbE6Il4FLiY1ewEsArYvO3y7nFYp/TlgiKQBHdLbnStv3zLv305ETImIsRExdtiwYet6uWZm611EP+tzyX0clwKPRMQ5ZenDy3b7EPBQfj0DOCqP9NoRGA38GZgNjM4jwwaSOv1nREQAs4Aj8vETgevLzjUxvz4CuC3vb2bWp7zySgowjVZzGdD9Lj22N/AJ4EFJ9+e0r5NGe+0CBPAY8GmAiJgraTrwMGmk2fERsRpA0gnATcCGwGURMTef7yTgaknfAu4jBTPy85WS5gOtpIBkZtbnNOJCYdCLwSUi/gh0NkLrxi6OmQxM7iT9xs6Oi4hHWdOsVp6+Evj3WsprZtaMGjW4+A59M7MmVgou/abPxczMet/KlenZNRczMyuMm8XMzKxwDi5mZlY497mYmVnh3OdiZmaFc7OYmZkVzsHFzMwK5+BiZmaFK/W5uEPfzMwK45qLmZkVzkORzcyscKW1XBptIXcHFzOzJrZyZePVWsDBxcysqTXiEsfg4GJm1tQcXMzMrHAOLmZmVjj3uZiZWeFcczEzs8I5uJiZWeEcXMzMrHDuczEzs8K55mJmZoVzcDEzs8I5uJiZWeFWruxnwUXS9pJmSXpY0lxJX8zpQyXNlDQvP2+V0yXpPEnzJT0gadeyc03M+8+TNLEsfTdJD+ZjzpPSvKCV8jAz60tWrUqP/tahvwr4ckSMAfYEjpc0BjgZuDUiRgO35p8BDgFG58ck4CJIgQI4DdgD2B04rSxYXAQcV3bcuJxeKQ8zsz6jURcKg14MLhGxOCLuza+XA48AI4DxwNS821Tg8Px6PHBFJHcBQyQNBw4GZkZEa0QsBWYC4/K2LSLirogI4IoO5+osDzOzPqNfBpdyklqAdwJ3A9tExOK86Slgm/x6BPBE2WELc1pX6Qs7SaeLPMzM+oyVK9NzvwwukjYHfgGcGBHLyrflGkf0Zv5d5SFpkqQ5kuYsWbKkN4thZla4Rl3iGHo5uEjaiBRYpkXEdTn56dykRX5+JqcvArYvO3y7nNZV+nadpHeVRzsRMSUixkbE2GHDhvXsIs3M6qRfNovlkVuXAo9ExDllm2YApRFfE4Hry9KPzqPG9gReyE1bNwEHSdoqd+QfBNyUty2TtGfO6+gO5+osDzOzPqORg8uAXjz33sAngAcl3Z/Tvg6cBUyXdCzwOPCRvO1G4FBgPtAGfBIgIlolnQnMzvt9MyJa8+vPAZcDmwK/zQ+6yMPMrM9o5D6XXgsuEfFHQBU279/J/gEcX+FclwGXdZI+B3h7J+nPdZaHmVlf0m/7XMzMrPc0crOYg4uZWZNycDEzs8I5uJiZWeEauUPfwcXMrEm5Q9/MzArnZjEzMyvcihUwcCBs0ICf5A1YJDMzq0ajLhQGDi5mZk1rxYrG7G8BBxczs6a1YoVrLmZmVrCmDy6SNuztgpiZWW36Qp/LPEnfkzSmV0tjZmZV6wt9Lu8A/g5cIumuvILjFr1YLjMz60bTN4tFxPKIuDgi9gJOAk4DFkuaKmlUr5bQzMw61fTBRdKGkj4o6ZfAD4DvA28Efk1a5MvMzNazRu5zqXaxsHnALOB7EXFnWfq1kt5bfLHMzKw7jdzn0m1wySPFLo+Ib3a2PSK+UHipzMysW03dLBYRq4HD1kNZzMysBo0cXKptFrtD0vnANcC/SokRcW+vlMrMzLrVF/pcdsnP5U1jAexXbHHMzKwaq1fDyy83bnCpdijyvp08HFjMzHpi2jRoaUlz5be0pJ9rVFqFsmk79EskvR/YCXjtUip18puZWQXTpsGkSdDWln5+/PH0M8CECVWfppEXCoMqg4ukHwGDgH2BS4AjgD/3YrnMzJrCnXfCuedCRJUH/HYraJvaPq0NmDQIfll9vn0iuAB7RcTOkh6IiDMkfR/4bW8WzMysGUydCr/6FbzlLVUe0DayQjrw19ry3nVX2H332o5ZX6oNLjlG0iZpW+A5YHjvFMnMrHm0tsKoUfDQQ1Ue0HJYagrraIcd4KHHiixaXVU7ceVvJA0BvgfcCzwGXNVbhTIzaxZLl8LQoTUcMHkyDBrUPm3QoJTeh1Q7WuzMiHg+In4B7AC8NSL+X1fHSLpM0jOSHipLO13SIkn358ehZdtOkTRf0t8kHVyWPi6nzZd0cln6jpLuzunXSBqY0zfOP8/P21uqfTPMzGrV2gpbbVXDARMmwJQpqaYipecpU2rqzG8GXTaLSfpwF9uIiOu6OPxy4Hzgig7p50bEf3c41xjgKNJotG2BWyS9OW++ADgQWAjMljQjIh4GvpvPdXUecHAscFF+XhoRoyQdlfc7sqvrNDPrqdZWePvbazxowoQ+F0w66q7P5QNdbAugYnCJiN/XUGsYD1wdES8B/5Q0Hyh1U82PiEcBJF0NjJf0COkGzo/lfaYCp5OCy/j8GuBa4HxJiqh6LIeZWdVqbhbrJ7oMLhHxyV7I8wRJRwNzgC9HxFJgBHBX2T4LcxrAEx3S9wBeBzwfEas62X9E6ZiIWCXphbz/sx0LImkSMAlg5MgKIzjMzCpYtQqWLauxWayfWN83UV4EnEmq9ZxJWhfmP2o8R2EiYgowBWDs2LGu2ZhZTZ5/Pj275rK2ahcL+xGp3+LzgIB/J3Xs1yQino6I1RHxKnAxa5q+FgHbl+26XU6rlP4cMETSgA7p7c6Vt2+Z9zczK1Rra3p2cFlbtUOR94qIo0kd5WcA7wbe3M0xa5FUfm/Mh4DSSLIZwFF5pNeOwGjSDACzgdF5ZNhAUqf/jNx/Mos0UwDAROD6snNNzK+PAG5zf4uZ9YZScHGz2Np6ehNlK93cRCnpKmAfYGtJC4HTgH0k7UJqFnsM+DRARMyVNB14GFgFHJ/XkUHSCcBNwIbAZRExN2dxEnC1pG8B9wGX5vRLgSvzoIBWUkAyMyvc0qXp2TWXtVUbXEo3UZ4N3JPTLunqgIj4aCfJl3aSVtp/MrDWXUQRcSNwYyfpj7KmWa08fSWp2c7MrFe5Wayy7u5zeRfwREScmX/eHHiQNAPOub1fPDOzxuVmscq663P5MfAygKT3AmfltBfIo6zMzPqrUrOYg8vaumsW2zAicmzmSGBKngLmF5Lu792imZk1ttZWGDwYBlR9U0f/0V3NZcOy4b77A7eVbfPbaWb9Wmur+1sq6S5AXAX8TtKzpBFjfwCQNIrUNGZm1m956pfKupv+ZbKkW0nDjm8uu19kA9INlWZm/VbNMyL3I902bUXEXZ2k/b13imNm1jyWLoWddqp3KRpTtXfom5lZB+5zqczBxcysByLcLNYVBxczsx5oa4NXXnHNpRIHFzOzHvDd+V1zcDEz6wHPK9Y1Bxczsx7wjMhdc3AxM+sBN4t1zcHFzKwH3CzWNQcXM7MecLNY1xxczMx6oLU1zYa82Wb1LkljcnAxM+uB0qSVUr1L0pgcXMzMesBTv3TNwcXMrAc89UvXHFzMzHrAa7l0zcHFzKwHXHPpmoOLmVkPuM+law4uZmY1WrUKli1zcOmKg4uZWY2efz49u1msMgcXM7Ma+e787vVacJF0maRnJD1UljZU0kxJ8/LzVjldks6TNF/SA5J2LTtmYt5/nqSJZem7SXowH3OelG5lqpSHmVlRPK9Y93qz5nI5MK5D2snArRExGrg1/wxwCDA6PyYBF0EKFMBpwB7A7sBpZcHiIuC4suPGdZOHmVkhPCNy93otuETE74HWDsnjgan59VTg8LL0KyK5CxgiaThwMDAzIlojYikwExiXt20REXdFRABXdDhXZ3mYmRXCzWLdW999LttExOL8+ilgm/x6BPBE2X4Lc1pX6Qs7Se8qj7VImiRpjqQ5S5Ys6cHlmFl/5Gax7tWtQz/XOKKeeUTElIgYGxFjhw0b1ptFMbM+pBRchgypbzka2foOLk/nJi3y8zM5fRGwfdl+2+W0rtK36yS9qzzMzAqxdCkMHgwbbVTvkjSu9R1cZgClEV8TgevL0o/Oo8b2BF7ITVs3AQdJ2ip35B8E3JS3LZO0Zx4ldnSHc3WWh5lZITz1S/cG9NaJJV0F7ANsLWkhadTXWcB0SccCjwMfybvfCBwKzAfagE8CRESrpDOB2Xm/b0ZEaZDA50gj0jYFfpsfdJGHmVkhPPVL95S6JWzs2LExZ86cehfDzHrTtGlw6qmwYAGMHAmTJ8OECTWf5j3vgY03hltv7YUyNhlJ90TE2I7pvkPfzPqHadNg0iR4/HGISM+TJqX0GrlZrHu91ixmZtabFi6E73wHXnmlygOuErSd2z6tDfiM4He15b1gQaq9WGUOLmbWlC6+GC68EIYPr/KAF/epkA78pra8t9wS9t23tmP6GwcXM2tKM2fC7rvD3XdXeUDLXqkprKMddoDHHiuyaIb7XMysCb3wAvz5z3DggTUcNHkyDBrUPm3QoJRuhXNwMbOmc/vtsHp1jcFlwgSYMiXVVKT0PGVKj0aLWffcLGZmTWfmzFTp2HPPGg+cMMHBZD1xzcXMms7MmfC+96V7TawxObiYWVNZsAD+/vcam8RsvXNwMbOmcsst6fmAA+pbDuuag4uZNZVbboE3vAHe/vZ6l8S64uBiZk3j1VdTcDnggDTgyxqXg4uZ9b5p06ClBTbYID33YD4vgAcegCVL3CTWDDwU2cx6V2nCyLa29HNpwkioeViw+1uah4OLmdVk3rw07cry5VUesPpI4Mj2aW3Ax1mzrF+1p1oNb3sbjBhR23G2/jm4mFlNLrwQXnwRvva11MrVrcnfBTpbN0pw8qk153/wwTUfYnXgxcIyLxZm1r0VK1Kt4YADYPr0Kg9qafGEkX2YFwszs3V27bWwdCl8+tM1HOQJI/slBxczq9qPfwyjRtW4loknjOyX3OdiZlWZOxfuuAPOPrvKvpZynjCy33HNxawvK+j+Eki1loED4Zhjiiqc9WWuuZj1VQXeX9LWBldcAR/+MAwbVnA5rU9ycDFrEhHw/e93PvCqU1Nfgbaz2qe1AZ99Be6qLe+FC9PqjzV15Fu/5uBi1iRuuw2++lXYYgsYUM1/7vIPVEgHflZ7/gcckNZQMauGg4tZk5g8GYYPh0cfhU02qeKAlt18f4nVjTv0zZrAnXfCrFmp5lJVYAHfX2J1VZfgIukxSQ9Kul/SnJw2VNJMSfPy81Y5XZLOkzRf0gOSdi07z8S8/zxJE8vSd8vnn5+P9eTc/UmBI6QaxeTJsPXWa/rjq+L7S6yO6llz2TcidimbNuBk4NaIGA3cmn8GOAQYnR+TgIsgBSPgNGAPYHfgtFJAyvscV3bcuN6/HGsIpRFSjz+eesBLI6TWd4ApMMDdey/ceCOceCJstlmNB0+YkJrAXn01PTuw2HrSSH0u44F98uupwO3ASTn9ikiToN0laYik4XnfmRHRCiBpJjBO0u3AFhFxV06/Ajgc+O16uxIrzCmnwA9+UMMBL/1fiA+3T2sDPiH4VG15DxwIhx8On/tcmgW46vpvgUOAAb79bdhySzjhhJoPNaubegWXAG6WFMCPI2IKsE1ELM7bnwK2ya9HAE+UHbswp3WVvrCT9LVImkSqDTFy5Mh1uR7rBX/6E5x1FowbBzvvXOVBZ59HpzPwhuALX6sp/yVL4Oc/T/d37LYb7L13lQHmspegrUO/Rhvw2Zdgdk1FYNUquO46+PrXU4Axaxb1Ci7viYhFkl4PzJT01/KNERE58PSqHNSmQJoVubfzs+qtXp1qDCNGpA/4zTev8sBrLqw8Quq7tQUXgB/+EH7603R3+tSpVR60/MMV0oHLay4Co0alJjGzZlKX4BIRi/LzM5J+SeozeVrS8IhYnJu9nsm7LwK2Lzt8u5y2iDXNaKX023P6dp3sb71t2jQ49VRYsABGjky90D1s47/oIrj//jSte9WBBVKe5U1SsE4jpAYPhs9+Nj2q1rKLhwBbv7feO/QlbSZpcOk1cBDwEDCDNevSTQSuz69nAKcsQm0AAAhqSURBVEfnUWN7Ai/k5rObgIMkbZU78g8CbsrblknaM48SO7rsXNZbCuxIf/pp+MY30k17RxxR48GNMELKQ4DN6lJz2Qb4ZR4dPAD4WUT8r6TZwHRJxwKPAx/J+98IHArMJ7VcfxIgIlolncmaVuxvljr3gc+RGiA2JXXkuzO/RsuXww03pDb/qpw4G9o+1D6tDfjibIjaPtinT08Vj/PPr6ETvVy9Z+At5V1QLc6sGXklyswrUa7x8sup1vCHP9SvDKefDqedVr/8zaw6lVaibKShyNYg/vM/U2CZMqWGRaHe9z54spOurW1HwO9+V1P+G22UWrPMrHk5uPQFBXakX3opXHABfOUrcNxxNRx49qTOO9LPPgNG9agoZtbEHFwa0IIFsGJFlTvPmAH/9T1YuTEwOvVWfep78ORg+OAHa8r3H/9Iw38PPBC+850aC+1+BjMr4z6XrFH6XG67Dfbfv375v/GNMHs2DB1avzKYWfNwn0uTuOGGNO3IT35S5Uipj32MTu9IR/Cz2hft2G8/BxYzW3cOLg1m1ix497tzzKjGKXdWvmHvo4UWzcysal7PpYG0tqa70qseoQW+Yc/MGpKDSwP5/e/Tze01BZdGuCPdzKwDN4s1kFmz0iqDe+xR44H1viPdzKwD11wayKxZaVr3jTeud0nMzNaNg0uDWLIEHnywxiYxM7MG5eDSIEozpDi4mFlf4ODSIGbNSuujv+td9S6Jmdm6c3BpELNmwXvekyZtNDNrdg4u62LaNGhpgQ02SM89WBgL4Kmn4JFH3CRmZn2HhyL3VGnlxdIswKWVF6HmYcG3356eHVzMrK9wcOmpU0+Ftja+xalcVZpnpQ345Ebw7dpOtWRJWqt9110LL6WZWV04uPTUggUAvIGnGMPDa9JfAca8uebT7b8/DPBvw8z6CE+5n9U85X5LS+UJIx97rKhimZk1tEpT7rtDv6c8YaSZWUUOLj3lCSPNzCpyK/+68ISRZmadcs3FzMwK5+BiZmaFc3AxM7PCObiYmVnhHFzMzKxwvokyk7QE6OSuyKayNfBsvQvRQPx+rOH3oj2/H+2ty/uxQ0QM65jo4NKHSJrT2Z2y/ZXfjzX8XrTn96O93ng/3CxmZmaFc3AxM7PCObj0LVPqXYAG4/djDb8X7fn9aK/w98N9LmZmVjjXXMzMrHAOLmZmVjgHlz5A0vaSZkl6WNJcSV+sd5nqTdKGku6T9Jt6l6XeJA2RdK2kv0p6RNK7612mepH0pfw/8pCkqyRtUu8yrU+SLpP0jKSHytKGSpopaV5+3qqIvBxc+oZVwJcjYgywJ3C8pDF1LlO9fRF4pN6FaBA/BP43It4KvIN++r5IGgF8ARgbEW8HNgSOqm+p1rvLgXEd0k4Gbo2I0cCt+ed15uDSB0TE4oi4N79eTvrwGFHfUtWPpO2A9wOX1Lss9SZpS+C9wKUAEfFyRDxf31LV1QBgU0kDgEHAk3Uuz3oVEb8HWjskjwem5tdTgcOLyMvBpY+R1AK8E7i7viWpqx8AXwNerXdBGsCOwBLgJ7mZ8BJJm9W7UPUQEYuA/wYWAIuBFyLi5vqWqiFsExGL8+ungG2KOKmDSx8iaXPgF8CJEbGs3uWpB0mHAc9ExD31LkuDGADsClwUEe8E/kVBzR7NJvcljCcF3G2BzSR9vL6laiyR7k0p5P4UB5c+QtJGpMAyLSKuq3d56mhv4IOSHgOuBvaT9NP6FqmuFgILI6JUk72WFGz6owOAf0bEkoh4BbgO2KvOZWoET0saDpCfnynipA4ufYAkkdrUH4mIc+pdnnqKiFMiYruIaCF11t4WEf3222lEPAU8IektOWl/4OE6FqmeFgB7ShqU/2f2p58ObuhgBjAxv54IXF/ESR1c+oa9gU+QvqXfnx+H1rtQ1jA+D0yT9ACwC/DtOpenLnLt7VrgXuBB0udfv5oGRtJVwJ+At0haKOlY4CzgQEnzSLW7swrJy9O/mJlZ0VxzMTOzwjm4mJlZ4RxczMyscA4uZmZWOAcXMzMrnIOLWTckrc7Dux+S9HNJg2o8fltJ1+bXu5QPE5f0QUmF3DEv6cUiztPb57T+wUORzboh6cWI2Dy/ngbc09ObVSUdQ5qV94QCi1g692vlbORzWv/gmotZbf4AjMprYPxK0gOS7pK0M4Ck95XdyHqfpMGSWnKtZyDwTeDIvP1IScdIOj8f2yLptnzOWyWNzOmXSzpP0p2SHpV0RHeFlPRVSbPzuc7IaWdJOr5sn9MlfaXS/mbrwsHFrEp5mvZDSHd3nwHcFxE7A18Hrsi7fQU4PiJ2Af4NWFE6PiJeBv4LuCYidomIazpk8T/A1HzOacB5ZduGA+8BDqObO6glHQSMBnYn3ZG/m6T3AtcAHynb9SPANV3sb9ZjDi5m3dtU0v3AHNL8VJeSPuivBIiI24DXSdoCuAM4R9IXgCERsaqGfN4N/Cy/vjLnUfKriHg1Ih6m+ynRD8qP+0hTnbwVGB0R9wGvz31A7wCWRsQTlfavodxmaxlQ7wKYNYEVuSbymjTv4doi4ixJNwCHAndIOhhYWUAZXirPvpt9BXwnIn7cybafA0cAbyDVZLrb36xHXHMx65k/ABMAJO0DPBsRyyS9KSIejIjvArNJtYByy4HBFc55J2uW3Z2Q8+iJm4D/yOv7IGmEpNfnbdfkPI4gBZru9jfrEddczHrmdOCyPNNwG2umLD9R0r6kVTDnAr8l9ZeUzAJOzs1s3+lwzs+TVoz8Kmn1yE/2pGARcbOktwF/yjWsF4GPkxZRmytpMLCotPpgV/v3JH8z8FBkMzPrBW4WMzOzwjm4mJlZ4RxczMyscA4uZmZWOAcXMzMrnIOLmZkVzsHFzMwK9/8BkcNzUntADsYAAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n",
        "---\n",
        "\n"
      ],
      "metadata": {
        "id": "wNCYyKwp34xu"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# IV. Random Forest Regression"
      ],
      "metadata": {
        "id": "d8Jt73IO34xw"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "v4S2fyIBYDcu"
      },
      "source": [
        "*Training the Random Forest Regression model on the whole dataset*"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o8dOCoJ1YKMc",
        "outputId": "af790c5f-d285-424c-ba33-28f44aa7ccc5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "from sklearn.ensemble import RandomForestRegressor\n",
        "my_RandomForestRegressor = RandomForestRegressor(n_estimators = 10, random_state = 0)\n",
        "my_RandomForestRegressor.fit(x, y)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:3: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n",
            "  This is separate from the ipykernel package so we can avoid doing imports until\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "RandomForestRegressor(n_estimators=10, random_state=0)"
            ]
          },
          "metadata": {},
          "execution_count": 177
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BzDueChsjmbu"
      },
      "source": [
        "*Predicting the output for a single value*"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pTXrS8FEYQlJ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d7203571-4efb-4c2a-a910-b0a097545e1a"
      },
      "source": [
        "my_RandomForestRegressor.predict([[5.5]])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ 49400])"
            ]
          },
          "metadata": {},
          "execution_count": 173
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kLqF9yMbYTon"
      },
      "source": [
        "*Visualising the Random Forest Regression results (higher resolution)*"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BMlTBifVYWNr",
        "outputId": "7cc70d4a-9b6e-4594-e871-fb8902abb30b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "x_grid = np.arange(min(x), max(x), 0.01)\n",
        "x_grid = x_grid.reshape((len(x_grid), 1))\n",
        "plt.scatter(x, y, color = 'red')\n",
        "plt.plot(x_grid, my_RandomForestRegressor.predict(x_grid), color = 'blue')\n",
        "plt.title('Truth or Bluff (Random Forest Regression)')\n",
        "plt.xlabel('Position level')\n",
        "plt.ylabel('Salary')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}