Binary_Class_Classification.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Binary Class Classification.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyMxZeQ8RmP5k8v3GOBsR/Bg",
      "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/Shuaib11-Github/Projects/blob/main/Binary_Class_Classification.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "n0eul-acgYf2"
      },
      "source": [
        "**Problem Definition**\n",
        "\n",
        "The focus of this project will be the Sonar Mines vs Rocks dataset1\n",
        "\n",
        ". The problem is to predict\n",
        "metal or rock objects from sonar return data. Each pattern is a set of 60 numbers in the range\n",
        "0.0 to 1.0. Each number represents the energy within a particular frequency band, integrated\n",
        "over a certain period of time. The label associated with each record contains the letter R if\n",
        "the object is a rock and M if it is a mine (metal cylinder). The numbers in the labels are in\n",
        "increasing order of aspect angle, but they do not encode the angle directly.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5j3LVIktgNPj"
      },
      "source": [
        "# Load libraries\n",
        "import numpy\n",
        "from matplotlib import pyplot\n",
        "from pandas import read_csv\n",
        "from pandas import set_option\n",
        "from pandas.plotting import scatter_matrix\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.model_selection import KFold\n",
        "from sklearn.model_selection import cross_val_score\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "from sklearn.metrics import classification_report\n",
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.pipeline import Pipeline\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
        "from sklearn.naive_bayes import GaussianNB\n",
        "from sklearn.svm import SVC\n",
        "from sklearn.ensemble import AdaBoostClassifier\n",
        "from sklearn.ensemble import GradientBoostingClassifier\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.ensemble import ExtraTreesClassifier\n"
      ],
      "execution_count": 41,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rigASgljgmA2"
      },
      "source": [
        "# Load dataset\n",
        "url = 'sonar.all-data.csv'\n",
        "dataset = read_csv(url, header=None)"
      ],
      "execution_count": 42,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "W_cYhS5Khs5O"
      },
      "source": [
        "You can see that we are not specifying the names of the attributes this time. This is because\n",
        "other than the class attribute (the last column), the variables do not have meaningful names.\n",
        "We also indicate that there is no header information, this is to avoid file loading code taking the\n",
        "first record as the column names. Now that we have the dataset loaded we can take a look at it."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h9yOUs69hwMK"
      },
      "source": [
        "**Descriptive Statistics**\n",
        "\n",
        "We will start off by confirming the dimensions of the dataset, e.g. the number of rows and\n",
        "columns."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yfv18SNLhl0Y",
        "outputId": "093ba8d6-63f1-4804-a782-46b93b94bed3"
      },
      "source": [
        "# shape\n",
        "print(dataset.shape)"
      ],
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(208, 61)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TmBvk3bjh5YP"
      },
      "source": [
        "We have 208 instances to work with and can confirm the data has 61 attributes including\n",
        "the class attribute.\n",
        "\n",
        "Let’s also look at the data types of each attribute."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ezdSU4kbhnuT",
        "outputId": "35058be9-9096-4157-81d8-d17df640ab62"
      },
      "source": [
        "# types\n",
        "set_option('display.max_rows', 500)\n",
        "print(dataset.dtypes)"
      ],
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0     float64\n",
            "1     float64\n",
            "2     float64\n",
            "3     float64\n",
            "4     float64\n",
            "5     float64\n",
            "6     float64\n",
            "7     float64\n",
            "8     float64\n",
            "9     float64\n",
            "10    float64\n",
            "11    float64\n",
            "12    float64\n",
            "13    float64\n",
            "14    float64\n",
            "15    float64\n",
            "16    float64\n",
            "17    float64\n",
            "18    float64\n",
            "19    float64\n",
            "20    float64\n",
            "21    float64\n",
            "22    float64\n",
            "23    float64\n",
            "24    float64\n",
            "25    float64\n",
            "26    float64\n",
            "27    float64\n",
            "28    float64\n",
            "29    float64\n",
            "30    float64\n",
            "31    float64\n",
            "32    float64\n",
            "33    float64\n",
            "34    float64\n",
            "35    float64\n",
            "36    float64\n",
            "37    float64\n",
            "38    float64\n",
            "39    float64\n",
            "40    float64\n",
            "41    float64\n",
            "42    float64\n",
            "43    float64\n",
            "44    float64\n",
            "45    float64\n",
            "46    float64\n",
            "47    float64\n",
            "48    float64\n",
            "49    float64\n",
            "50    float64\n",
            "51    float64\n",
            "52    float64\n",
            "53    float64\n",
            "54    float64\n",
            "55    float64\n",
            "56    float64\n",
            "57    float64\n",
            "58    float64\n",
            "59    float64\n",
            "60     object\n",
            "dtype: object\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KYTP5r5iiJXI"
      },
      "source": [
        "We can see that all of the attributes are numeric (float) and that the class value has been\n",
        "read in as an object.\n",
        "\n",
        "Let’s now take a peek at the first 20 rows of the data.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aL-91I6ciBXF",
        "outputId": "c07ac885-6087-4e06-c8ce-3057114b76a8"
      },
      "source": [
        "# head\n",
        "set_option('display.width', 100)\n",
        "print(dataset.head(20))"
      ],
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "       0      1      2      3      4   ...     56         57         58     59  60\n",
            "0   0.020  0.037  0.043  0.021  0.095  ...  0.018  8.400e-03  9.000e-03  0.003   R\n",
            "1   0.045  0.052  0.084  0.069  0.118  ...  0.014  4.900e-03  5.200e-03  0.004   R\n",
            "2   0.026  0.058  0.110  0.108  0.097  ...  0.032  1.640e-02  9.500e-03  0.008   R\n",
            "3   0.010  0.017  0.062  0.021  0.021  ...  0.005  4.400e-03  4.000e-03  0.012   R\n",
            "4   0.076  0.067  0.048  0.039  0.059  ...  0.007  4.800e-03  1.070e-02  0.009   R\n",
            "5   0.029  0.045  0.028  0.017  0.038  ...  0.006  2.700e-03  5.100e-03  0.006   R\n",
            "6   0.032  0.096  0.132  0.141  0.167  ...  0.009  1.430e-02  3.600e-03  0.010   R\n",
            "7   0.052  0.055  0.084  0.032  0.116  ...  0.009  4.700e-03  4.800e-03  0.005   R\n",
            "8   0.022  0.037  0.048  0.048  0.065  ...  0.006  9.300e-03  5.900e-03  0.002   R\n",
            "9   0.016  0.017  0.035  0.007  0.019  ...  0.003  3.500e-03  5.600e-03  0.004   R\n",
            "10  0.004  0.006  0.015  0.034  0.031  ...  0.004  3.000e-04  5.300e-03  0.004   R\n",
            "11  0.012  0.031  0.017  0.031  0.036  ...  0.003  9.200e-03  9.000e-04  0.004   R\n",
            "12  0.008  0.009  0.005  0.025  0.034  ...  0.006  5.800e-03  5.900e-03  0.003   R\n",
            "13  0.009  0.006  0.025  0.049  0.120  ...  0.016  5.300e-03  1.890e-02  0.010   R\n",
            "14  0.012  0.043  0.060  0.045  0.060  ...  0.011  1.960e-02  1.470e-02  0.006   R\n",
            "15  0.030  0.061  0.065  0.092  0.162  ...  0.015  4.900e-03  2.000e-02  0.007   R\n",
            "16  0.035  0.012  0.019  0.047  0.074  ...  0.009  1.500e-03  7.300e-03  0.007   R\n",
            "17  0.019  0.061  0.038  0.077  0.139  ...  0.004  3.700e-03  1.120e-02  0.007   R\n",
            "18  0.027  0.009  0.015  0.028  0.041  ...  0.012  1.320e-02  7.000e-03  0.009   R\n",
            "19  0.013  0.015  0.064  0.173  0.257  ...  0.018  9.400e-03  1.160e-02  0.006   R\n",
            "\n",
            "[20 rows x 61 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WVVIn9K7iXkj"
      },
      "source": [
        "This does not show all of the columns, but we can see all of the data has the same scale. We\n",
        "can also see that the class attribute (60) has string values."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "s_mlxXjgibeY"
      },
      "source": [
        "Let’s summarize the distribution of each attribute.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VEoHVrLDiUY0",
        "outputId": "6213dfea-2b51-4c19-a2b6-87ff755a914e"
      },
      "source": [
        "# descriptions, change precision to 3 places\n",
        "set_option('precision', 3)\n",
        "print(dataset.describe())"
      ],
      "execution_count": 46,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "            0          1        2   ...         57         58         59\n",
            "count  208.000  2.080e+02  208.000  ...  2.080e+02  2.080e+02  2.080e+02\n",
            "mean     0.029  3.844e-02    0.044  ...  7.949e-03  7.941e-03  6.507e-03\n",
            "std      0.023  3.296e-02    0.038  ...  6.470e-03  6.181e-03  5.031e-03\n",
            "min      0.002  6.000e-04    0.002  ...  3.000e-04  1.000e-04  6.000e-04\n",
            "25%      0.013  1.645e-02    0.019  ...  3.600e-03  3.675e-03  3.100e-03\n",
            "50%      0.023  3.080e-02    0.034  ...  5.800e-03  6.400e-03  5.300e-03\n",
            "75%      0.036  4.795e-02    0.058  ...  1.035e-02  1.033e-02  8.525e-03\n",
            "max      0.137  2.339e-01    0.306  ...  4.400e-02  3.640e-02  4.390e-02\n",
            "\n",
            "[8 rows x 60 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A4PsKKmCihxW"
      },
      "source": [
        "Again, as we expect, the data has the same range, but interestingly differing mean values.\n",
        "There may be some benefit from standardizing the data.\n",
        "\n",
        "Let’s take a quick look at the breakdown of class values."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4q5I-yALieUU",
        "outputId": "63e20969-9c15-4422-89f3-d19853aaff0c"
      },
      "source": [
        "# class distribution\n",
        "print(dataset.groupby(60).size())"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "60\n",
            "M    111\n",
            "R     97\n",
            "dtype: int64\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3tKE6u44isnH"
      },
      "source": [
        "We can see that the classes are reasonably balanced between M (mines) and R (rocks)."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sx6VBVnniwFj"
      },
      "source": [
        "**Unimodal Data Visualizations**\n",
        "\n",
        "Let’s look at visualizations of individual attributes. It is often useful to look at your data\n",
        "using multiple different visualizations in order to spark ideas. Let’s look at histograms of each\n",
        "attribute to get a sense of the data distributions."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 598
        },
        "id": "x1l-k7m9ioxE",
        "outputId": "512a71b4-a036-49ea-9a44-7fde368815af"
      },
      "source": [
        "# histograms\n",
        "dataset.hist(sharex=False, sharey=False, figsize = (20,10), xlabelsize=1, ylabelsize=1)\n",
        "pyplot.show()"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x720 with 64 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OHBW1hzZi7jE"
      },
      "source": [
        "We can see that there are a lot of Gaussian-like distributions and perhaps some exponential like distributions for other attributes\n",
        "\n",
        "Let’s take a look at the same perspective of the data using density plots."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 583
        },
        "id": "53yQmOsZi22w",
        "outputId": "3a55cf73-d118-439a-c50e-42389e6d2c77"
      },
      "source": [
        "# density\n",
        "dataset.plot(kind='density', subplots=True, layout=(8,8), figsize = (20,10), sharex=False, legend=False,\n",
        "fontsize=1)\n",
        "pyplot.show()"
      ],
      "execution_count": 49,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x720 with 64 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UngX0mmPjnRs"
      },
      "source": [
        "We can see that attributes do have quite different spreads. Given the scales are the same, it\n",
        "may suggest some benefit in standardizing the data for modeling to get all of the means lined\n",
        "up."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qhdJCUuCkA14"
      },
      "source": [
        "**Multimodal Data Visualizations**\n",
        "\n",
        "Let’s visualize the correlations between the attributes."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 584
        },
        "id": "VZSmi73RjjLv",
        "outputId": "cf83f8e3-6d07-4765-8752-166e2205a419"
      },
      "source": [
        "# correlation matrix\n",
        "fig = pyplot.figure(figsize = (10, 10))\n",
        "ax = fig.add_subplot(111)\n",
        "cax = ax.matshow(dataset.corr(), vmin=-1, vmax=1, interpolation='none')\n",
        "fig.colorbar(cax)\n",
        "pyplot.show()"
      ],
      "execution_count": 50,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bgetpozikNKr"
      },
      "source": [
        "It looks like there is also some structure in the order of the attributes. The yellow around\n",
        "the diagonal suggests that attributes that are next to each other are generally more correlated\n",
        "with each other. The blue patches also suggest some moderate negative correlation the further\n",
        "attributes are away from each other in the ordering. This makes sense if the order of the\n",
        "attributes refers to the angle of sensors for the sonar chirp."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xsRwp0Gfka9-"
      },
      "source": [
        "**Validation Dataset**\n",
        "\n",
        "It is a good idea to use a validation hold-out set. This is a sample of the data that we hold\n",
        "back from our analysis and modeling. We use it right at the end of our project to confirm the\n",
        "accuracy of our final model. It is a smoke test that we can use to see if we messed up and to\n",
        "give us confidence on our estimates of accuracy on unseen data. We will use 80% of the dataset\n",
        "for modeling and hold back 20% for validation.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3SpOByX1kFLx"
      },
      "source": [
        "# Split-out validation dataset\n",
        "array = dataset.values\n",
        "X = array[:,0:60].astype(float)\n",
        "Y = array[:,60]\n",
        "validation_size = 0.20\n",
        "seed = 7\n",
        "X_train, X_validation, Y_train, Y_validation = train_test_split(X, Y,\n",
        "test_size=validation_size, random_state=seed)"
      ],
      "execution_count": 51,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nACtKu0Vkhfq"
      },
      "source": [
        "**Evaluate Algorithms: Baseline**\n",
        "\n",
        "We don’t know what algorithms will do well on this dataset. Gut feel suggests distance based\n",
        "algorithms like k-Nearest Neighbors and Support Vector Machines may do well. Let’s design\n",
        "our test harness. We will use 10-fold cross validation. The dataset is not too small and this is\n",
        "a good standard test harness configuration. We will evaluate algorithms using the accuracy\n",
        "metric. This is a gross metric that will give a quick idea of how correct a given model is. More\n",
        "useful on binary classification problems like this one."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "e4x196nnkemT"
      },
      "source": [
        "# Test options and evaluation metric\n",
        "num_folds = 10\n",
        "seed = 7\n",
        "scoring = 'accuracy'"
      ],
      "execution_count": 52,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tgVnb_W2koZ7"
      },
      "source": [
        "Let’s create a baseline of performance on this problem and spot-check a number of different\n",
        "algorithms. We will select a suite of different algorithms capable of working on this classification\n",
        "problem. The six algorithms selected include:\n",
        "\n",
        "1. **Linear Algorithms**: Logistic Regression (LR) and Linear Discriminant Analysis (LDA).\n",
        "2. **Nonlinear Algorithms**: Classification and Regression Trees (CART), Support Vector\n",
        "Machines (SVM), Gaussian Naive Bayes (NB) and k-Nearest Neighbors (KNN).\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tiW2SG4Dklrx"
      },
      "source": [
        "# Spot-Check Algorithms\n",
        "models = []\n",
        "models.append(('LR', LogisticRegression()))\n",
        "models.append(('LDA', LinearDiscriminantAnalysis()))\n",
        "models.append(('KNN', KNeighborsClassifier()))\n",
        "models.append(('CART', DecisionTreeClassifier()))\n",
        "models.append(('NB', GaussianNB()))\n",
        "models.append(('SVM', SVC()))"
      ],
      "execution_count": 53,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "J-gk2HsXkz6c"
      },
      "source": [
        "The algorithms all use default tuning parameters. Let’s compare the algorithms. We will\n",
        "display the mean and standard deviation of accuracy for each algorithm as we calculate it and\n",
        "collect the results for use later."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1aJIESiikw-e",
        "outputId": "422b3151-c708-4d74-8376-8065526aca99"
      },
      "source": [
        "results = []\n",
        "names = []\n",
        "for name, model in models:\n",
        "  kfold = KFold(n_splits=num_folds, random_state=seed)\n",
        "  cv_results = cross_val_score(model, X_train, Y_train, cv=kfold, scoring=scoring)\n",
        "  results.append(cv_results)\n",
        "  names.append(name)\n",
        "  msg = \"%s: %f (%f)\" % (name, cv_results.mean(), cv_results.std())\n",
        "  print(msg)"
      ],
      "execution_count": 54,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "LR: 0.769485 (0.100510)\n",
            "LDA: 0.746324 (0.117854)\n",
            "KNN: 0.808088 (0.067507)\n",
            "CART: 0.711765 (0.115376)\n",
            "NB: 0.648897 (0.141868)\n",
            "SVM: 0.776471 (0.090411)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "m5Nk9_oKlCVD"
      },
      "source": [
        "Running the example provides the output below. The results suggest That both Logistic\n",
        "Regression and k-Nearest Neighbors may be worth further study.\n",
        "\n",
        "LR: 0.769485 (0.100510)\n",
        "\n",
        "LDA: 0.746324 (0.117854)\n",
        "\n",
        "KNN: 0.808088 (0.067507)\n",
        "\n",
        "CART: 0.735662 (0.113477)\n",
        "\n",
        "NB: 0.648897 (0.141868)\n",
        "\n",
        "SVM: 0.776471 (0.090411)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uctETXizlRhy"
      },
      "source": [
        "These are just mean accuracy values. It is always wise to look at the distribution of accuracy\n",
        "values calculated across cross validation folds. We can do that graphically using box and whisker\n",
        "plots.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 294
        },
        "id": "AY_vNaklk8Vr",
        "outputId": "6d45bde6-9368-4037-d83f-f269a16f3566"
      },
      "source": [
        "# Compare Algorithms\n",
        "fig = pyplot.figure()\n",
        "fig.suptitle('Algorithm Comparison')\n",
        "ax = fig.add_subplot(111)\n",
        "pyplot.boxplot(results)\n",
        "ax.set_xticklabels(names)\n",
        "pyplot.show()"
      ],
      "execution_count": 55,
      "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"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e9Huvx55lXXD"
      },
      "source": [
        "The results show a tight distribution for KNN which is encouraging, suggesting low variance.\n",
        "The poor results for SVM are surprising."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WOXX4q2glcQ1"
      },
      "source": [
        "It is possible that the varied distribution of the attributes is having an effect on the accuracy\n",
        "of algorithms such as SVM. In the next section we will repeat this spot-check with a standardized\n",
        "copy of the training dataset.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bfND7kcplfr4"
      },
      "source": [
        "**Evaluate Algorithms: Standardize Data**\n",
        "\n",
        "We suspect that the differing distributions of the raw data may be negatively impacting the skill\n",
        "of some of the algorithms. Let’s evaluate the same algorithms with a standardized copy of the\n",
        "dataset. This is where the data is transformed such that each attribute has a mean value of zero\n",
        "and a standard deviation of one. We also need to avoid data leakage when we transform the\n",
        "data. A good way to avoid leakage is to use pipelines that standardize the data and build the\n",
        "model for each fold in the cross validation test harness. That way we can get a fair estimation\n",
        "of how each model with standardized data might perform on unseen data.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HpKqj_GTlUI5",
        "outputId": "2845b87b-117c-4bcb-bd0c-6d5f09d3065a"
      },
      "source": [
        "# Standardize the dataset\n",
        "pipelines = []\n",
        "pipelines.append(('ScaledLR', Pipeline([('Scaler', StandardScaler()),('LR',\n",
        "LogisticRegression())])))\n",
        "pipelines.append(('ScaledLDA', Pipeline([('Scaler', StandardScaler()),('LDA',\n",
        "LinearDiscriminantAnalysis())])))\n",
        "pipelines.append(('ScaledKNN', Pipeline([('Scaler', StandardScaler()),('KNN',\n",
        "KNeighborsClassifier())])))\n",
        "pipelines.append(('ScaledCART', Pipeline([('Scaler', StandardScaler()),('CART',\n",
        "DecisionTreeClassifier())])))\n",
        "pipelines.append(('ScaledNB', Pipeline([('Scaler', StandardScaler()),('NB',\n",
        "GaussianNB())])))\n",
        "pipelines.append(('ScaledSVM', Pipeline([('Scaler', StandardScaler()),('SVM', SVC())])))\n",
        "results = []\n",
        "names = []\n",
        "for name, model in pipelines:\n",
        "  kfold = KFold(n_splits=num_folds, random_state=seed)\n",
        "  cv_results = cross_val_score(model, X_train, Y_train, cv=kfold, scoring=scoring)\n",
        "  results.append(cv_results)\n",
        "  names.append(name)\n",
        "  msg = \"%s: %f (%f)\" % (name, cv_results.mean(), cv_results.std())\n",
        "  print(msg)"
      ],
      "execution_count": 56,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "ScaledLR: 0.740441 (0.094668)\n",
            "ScaledLDA: 0.746324 (0.117854)\n",
            "ScaledKNN: 0.825735 (0.054511)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "ScaledCART: 0.735662 (0.131517)\n",
            "ScaledNB: 0.648897 (0.141868)\n",
            "ScaledSVM: 0.836397 (0.088697)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0bBF5vmZluDM"
      },
      "source": [
        "Running the example provides the results listed below. We can see that KNN is still doing\n",
        "well, even better than before. We can also see that the standardization of the data has lifted\n",
        "the skill of SVM to be the most accurate algorithm tested so far.\n",
        "\n",
        "ScaledLR: 0.740441 (0.094668)\n",
        "\n",
        "ScaledLDA: 0.746324 (0.117854)\n",
        "\n",
        "ScaledKNN: 0.825735 (0.054511)\n",
        "\n",
        "ScaledCART: 0.776471 (0.083478)\n",
        "\n",
        "ScaledNB: 0.648897 (0.141868)\n",
        "\n",
        "ScaledSVM: 0.836397 (0.088697)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QgrOrKQVl9n0"
      },
      "source": [
        "Again, we should plot the distribution of the accuracy scores using box and whisker plots."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 294
        },
        "id": "QDJR9NrtlqhG",
        "outputId": "fafd4188-7d52-43d5-c32c-480fe32668d5"
      },
      "source": [
        "# Compare Algorithms\n",
        "fig = pyplot.figure()\n",
        "fig.suptitle('Scaled Algorithm Comparison')\n",
        "ax = fig.add_subplot(111)\n",
        "pyplot.boxplot(results)\n",
        "ax.set_xticklabels(names)\n",
        "pyplot.show()"
      ],
      "execution_count": 57,
      "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"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "V0FERi8LmHJH"
      },
      "source": [
        "The results suggest digging deeper into the SVM and KNN algorithms. It is very likely that\n",
        "configuration beyond the default may yield even more accurate models."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uYanNVwrmKkV"
      },
      "source": [
        "**Algorithm Tuning**\n",
        "\n",
        "In this section we investigate tuning the parameters for two algorithms that show promise from\n",
        "the spot-checking in the previous section: KNN and SVM.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "z4jbfo43mQ_4"
      },
      "source": [
        "**Tuning KNN**\n",
        "\n",
        "We can start off by tuning the number of neighbors for KNN. The default number of neighbors\n",
        "is 7. Below we try all odd values of k from 1 to 21, covering the default value of 7. Each k value\n",
        "is evaluated using 10-fold cross validation on the training standardized dataset."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PJvp8d2BmCzN",
        "outputId": "88e39f30-89e8-4a74-a6b8-71efada79ec0"
      },
      "source": [
        "# Tune scaled KNN\n",
        "scaler = StandardScaler().fit(X_train)\n",
        "rescaledX = scaler.transform(X_train)\n",
        "neighbors = [1,3,5,7,9,11,13,15,17,19,21]\n",
        "param_grid = dict(n_neighbors=neighbors)\n",
        "model = KNeighborsClassifier()\n",
        "kfold = KFold(n_splits=num_folds, random_state=seed)\n",
        "grid = GridSearchCV(estimator=model, param_grid=param_grid, scoring=scoring, cv=kfold)\n",
        "grid_result = grid.fit(rescaledX, Y_train)\n",
        "print(\"Best: %f using %s\" % (grid_result.best_score_, grid_result.best_params_))\n",
        "means = grid_result.cv_results_['mean_test_score']\n",
        "stds = grid_result.cv_results_['std_test_score']\n",
        "params = grid_result.cv_results_['params']\n",
        "for mean, stdev, param in zip(means, stds, params):\n",
        "  print(\"%f (%f) with: %r\" % (mean, stdev, param))"
      ],
      "execution_count": 58,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Best: 0.850000 using {'n_neighbors': 1}\n",
            "0.850000 (0.059686) with: {'n_neighbors': 1}\n",
            "0.837132 (0.066014) with: {'n_neighbors': 3}\n",
            "0.837500 (0.037377) with: {'n_neighbors': 5}\n",
            "0.763971 (0.089374) with: {'n_neighbors': 7}\n",
            "0.751471 (0.087051) with: {'n_neighbors': 9}\n",
            "0.733456 (0.104831) with: {'n_neighbors': 11}\n",
            "0.733088 (0.105806) with: {'n_neighbors': 13}\n",
            "0.727941 (0.076148) with: {'n_neighbors': 15}\n",
            "0.709926 (0.079287) with: {'n_neighbors': 17}\n",
            "0.722059 (0.085088) with: {'n_neighbors': 19}\n",
            "0.710294 (0.109505) with: {'n_neighbors': 21}\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5piHqE5VmePq"
      },
      "source": [
        "We can print out configuration that resulted in the highest accuracy as well as the accuracy\n",
        "of all values tried. Running the example we see the results below.\n",
        "\n",
        "Best: 0.850000 using {'n_neighbors': 1}\n",
        "\n",
        "0.850000 (0.059686) with: {'n_neighbors': 1}\n",
        "\n",
        "0.837132 (0.066014) with: {'n_neighbors': 3}\n",
        "\n",
        "0.837500 (0.037377) with: {'n_neighbors': 5}\n",
        "\n",
        "0.763971 (0.089374) with: {'n_neighbors': 7}\n",
        "\n",
        "0.751471 (0.087051) with: {'n_neighbors': 9}\n",
        "\n",
        "0.733456 (0.104831) with: {'n_neighbors': 11}\n",
        "\n",
        "0.733088 (0.105806) with: {'n_neighbors': 13}\n",
        "\n",
        "0.727941 (0.076148) with: {'n_neighbors': 15}\n",
        "\n",
        "0.709926 (0.079287) with: {'n_neighbors': 17}\n",
        "\n",
        "0.722059 (0.085088) with: {'n_neighbors': 19}\n",
        "\n",
        "0.710294 (0.109505) with: {'n_neighbors': 21}"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qC94guKvmvaS"
      },
      "source": [
        "We can see that the optimal configuration is K=1. This is interesting as the algorithm will\n",
        "make predictions using the most similar instance in the training dataset alone."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "S30Su2_0nFhr"
      },
      "source": [
        "**Tuning SVM**\n",
        "\n",
        "We can tune two key parameters of the SVM algorithm, the value of C (how much to relax the\n",
        "margin) and the type of kernel. The default for SVM (the SVC class) is to use the Radial\n",
        "Basis Function (RBF) kernel with a C value set to 1.0. Like with KNN, we will perform a grid\n",
        "search using 10-fold cross validation with a standardized copy of the training dataset. We will\n",
        "try a number of simpler kernel types and C values with less bias and more bias (less than and\n",
        "more than 1.0 respectively).\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "s_2gJ4jdmarP",
        "outputId": "03627ec0-cb82-4cc8-bd50-6750b9fc2279"
      },
      "source": [
        "# Tune scaled SVM\n",
        "scaler = StandardScaler().fit(X_train)\n",
        "rescaledX = scaler.transform(X_train)\n",
        "c_values = [0.1, 0.3, 0.5, 0.7, 0.9, 1.0, 1.3, 1.5, 1.7, 2.0]\n",
        "kernel_values = ['linear', 'poly', 'rbf', 'sigmoid']\n",
        "param_grid = dict(C=c_values, kernel=kernel_values)\n",
        "model = SVC()\n",
        "kfold = KFold(n_splits=num_folds, random_state=seed)\n",
        "grid = GridSearchCV(estimator=model, param_grid=param_grid, scoring=scoring, cv=kfold)\n",
        "grid_result = grid.fit(rescaledX, Y_train)\n",
        "print(\"Best: %f using %s\" % (grid_result.best_score_, grid_result.best_params_))\n",
        "means = grid_result.cv_results_['mean_test_score']\n",
        "stds = grid_result.cv_results_['std_test_score']\n",
        "params = grid_result.cv_results_['params']\n",
        "for mean, stdev, param in zip(means, stds, params):\n",
        "  print(\"%f (%f) with: %r\" % (mean, stdev, param))"
      ],
      "execution_count": 59,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Best: 0.872426 using {'C': 2.0, 'kernel': 'rbf'}\n",
            "0.758456 (0.099483) with: {'C': 0.1, 'kernel': 'linear'}\n",
            "0.535294 (0.114222) with: {'C': 0.1, 'kernel': 'poly'}\n",
            "0.573162 (0.130930) with: {'C': 0.1, 'kernel': 'rbf'}\n",
            "0.704044 (0.066933) with: {'C': 0.1, 'kernel': 'sigmoid'}\n",
            "0.746324 (0.109507) with: {'C': 0.3, 'kernel': 'linear'}\n",
            "0.642647 (0.132187) with: {'C': 0.3, 'kernel': 'poly'}\n",
            "0.765809 (0.091692) with: {'C': 0.3, 'kernel': 'rbf'}\n",
            "0.734926 (0.054567) with: {'C': 0.3, 'kernel': 'sigmoid'}\n",
            "0.740074 (0.082636) with: {'C': 0.5, 'kernel': 'linear'}\n",
            "0.662132 (0.105668) with: {'C': 0.5, 'kernel': 'poly'}\n",
            "0.788235 (0.064190) with: {'C': 0.5, 'kernel': 'rbf'}\n",
            "0.746691 (0.058742) with: {'C': 0.5, 'kernel': 'sigmoid'}\n",
            "0.746691 (0.084198) with: {'C': 0.7, 'kernel': 'linear'}\n",
            "0.746324 (0.129754) with: {'C': 0.7, 'kernel': 'poly'}\n",
            "0.812500 (0.085513) with: {'C': 0.7, 'kernel': 'rbf'}\n",
            "0.758824 (0.047088) with: {'C': 0.7, 'kernel': 'sigmoid'}\n",
            "0.758824 (0.096520) with: {'C': 0.9, 'kernel': 'linear'}\n",
            "0.769853 (0.111321) with: {'C': 0.9, 'kernel': 'poly'}\n",
            "0.836397 (0.088697) with: {'C': 0.9, 'kernel': 'rbf'}\n",
            "0.752941 (0.074130) with: {'C': 0.9, 'kernel': 'sigmoid'}\n",
            "0.752574 (0.098883) with: {'C': 1.0, 'kernel': 'linear'}\n",
            "0.781985 (0.101194) with: {'C': 1.0, 'kernel': 'poly'}\n",
            "0.836397 (0.088697) with: {'C': 1.0, 'kernel': 'rbf'}\n",
            "0.752574 (0.070872) with: {'C': 1.0, 'kernel': 'sigmoid'}\n",
            "0.769853 (0.106086) with: {'C': 1.3, 'kernel': 'linear'}\n",
            "0.818382 (0.107151) with: {'C': 1.3, 'kernel': 'poly'}\n",
            "0.848162 (0.080414) with: {'C': 1.3, 'kernel': 'rbf'}\n",
            "0.722794 (0.072066) with: {'C': 1.3, 'kernel': 'sigmoid'}\n",
            "0.758088 (0.092026) with: {'C': 1.5, 'kernel': 'linear'}\n",
            "0.830147 (0.110255) with: {'C': 1.5, 'kernel': 'poly'}\n",
            "0.866176 (0.091458) with: {'C': 1.5, 'kernel': 'rbf'}\n",
            "0.728309 (0.078171) with: {'C': 1.5, 'kernel': 'sigmoid'}\n",
            "0.746324 (0.090414) with: {'C': 1.7, 'kernel': 'linear'}\n",
            "0.830515 (0.116706) with: {'C': 1.7, 'kernel': 'poly'}\n",
            "0.860294 (0.088281) with: {'C': 1.7, 'kernel': 'rbf'}\n",
            "0.715441 (0.089652) with: {'C': 1.7, 'kernel': 'sigmoid'}\n",
            "0.758456 (0.094064) with: {'C': 2.0, 'kernel': 'linear'}\n",
            "0.824632 (0.118381) with: {'C': 2.0, 'kernel': 'poly'}\n",
            "0.872426 (0.097562) with: {'C': 2.0, 'kernel': 'rbf'}\n",
            "0.733456 (0.090179) with: {'C': 2.0, 'kernel': 'sigmoid'}\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7RGomXoJnTAB"
      },
      "source": [
        "Running the example prints out the best configuration, the accuracy as well as the accuracies\n",
        "for all configuration combinations.\n",
        "\n",
        "Best: 0.872426 using {'C': 2.0, 'kernel': 'rbf'}\n",
        "\n",
        "0.758456 (0.099483) with: {'C': 0.1, 'kernel': 'linear'}\n",
        "\n",
        "0.535294 (0.114222) with: {'C': 0.1, 'kernel': 'poly'}\n",
        "\n",
        "0.573162 (0.130930) with: {'C': 0.1, 'kernel': 'rbf'}\n",
        "\n",
        "0.704044 (0.066933) with: {'C': 0.1, 'kernel': 'sigmoid'}\n",
        "\n",
        "0.746324 (0.109507) with: {'C': 0.3, 'kernel': 'linear'}\n",
        "\n",
        "0.642647 (0.132187) with: {'C': 0.3, 'kernel': 'poly'}\n",
        "\n",
        "0.765809 (0.091692) with: {'C': 0.3, 'kernel': 'rbf'}\n",
        "\n",
        "0.734926 (0.054567) with: {'C': 0.3, 'kernel': 'sigmoid'}\n",
        "\n",
        "0.740074 (0.082636) with: {'C': 0.5, 'kernel': 'linear'}\n",
        "\n",
        "0.662132 (0.105668) with: {'C': 0.5, 'kernel': 'poly'}\n",
        "\n",
        "0.788235 (0.064190) with: {'C': 0.5, 'kernel': 'rbf'}\n",
        "\n",
        "0.746691 (0.058742) with: {'C': 0.5, 'kernel': 'sigmoid'}\n",
        "\n",
        "0.746691 (0.084198) with: {'C': 0.7, 'kernel': 'linear'}\n",
        "\n",
        "0.746324 (0.129754) with: {'C': 0.7, 'kernel': 'poly'}\n",
        "\n",
        "0.812500 (0.085513) with: {'C': 0.7, 'kernel': 'rbf'}\n",
        "\n",
        "0.758824 (0.047088) with: {'C': 0.7, 'kernel': 'sigmoid'}\n",
        "\n",
        "0.758824 (0.096520) with: {'C': 0.9, 'kernel': 'linear'}\n",
        "\n",
        "0.769853 (0.111321) with: {'C': 0.9, 'kernel': 'poly'}\n",
        "\n",
        "0.836397 (0.088697) with: {'C': 0.9, 'kernel': 'rbf'}\n",
        "\n",
        "0.752941 (0.074130) with: {'C': 0.9, 'kernel': 'sigmoid'}\n",
        "\n",
        "0.752574 (0.098883) with: {'C': 1.0, 'kernel': 'linear'}\n",
        "\n",
        "0.781985 (0.101194) with: {'C': 1.0, 'kernel': 'poly'}\n",
        "\n",
        "0.836397 (0.088697) with: {'C': 1.0, 'kernel': 'rbf'}\n",
        "\n",
        "0.752574 (0.070872) with: {'C': 1.0, 'kernel': 'sigmoid'}\n",
        "\n",
        "0.769853 (0.106086) with: {'C': 1.3, 'kernel': 'linear'}\n",
        "\n",
        "0.818382 (0.107151) with: {'C': 1.3, 'kernel': 'poly'}\n",
        "\n",
        "0.848162 (0.080414) with: {'C': 1.3, 'kernel': 'rbf'}\n",
        "\n",
        "0.722794 (0.072066) with: {'C': 1.3, 'kernel': 'sigmoid'}\n",
        "\n",
        "0.758088 (0.092026) with: {'C': 1.5, 'kernel': 'linear'}\n",
        "\n",
        "0.830147 (0.110255) with: {'C': 1.5, 'kernel': 'poly'}\n",
        "\n",
        "0.866176 (0.091458) with: {'C': 1.5, 'kernel': 'rbf'}\n",
        "\n",
        "0.728309 (0.078171) with: {'C': 1.5, 'kernel': 'sigmoid'}\n",
        "\n",
        "0.746324 (0.090414) with: {'C': 1.7, 'kernel': 'linear'}\n",
        "\n",
        "0.830515 (0.116706) with: {'C': 1.7, 'kernel': 'poly'}\n",
        "\n",
        "0.860294 (0.088281) with: {'C': 1.7, 'kernel': 'rbf'}\n",
        "\n",
        "0.715441 (0.089652) with: {'C': 1.7, 'kernel': 'sigmoid'}\n",
        "\n",
        "0.758456 (0.094064) with: {'C': 2.0, 'kernel': 'linear'}\n",
        "\n",
        "0.824632 (0.118381) with: {'C': 2.0, 'kernel': 'poly'}\n",
        "\n",
        "0.872426 (0.097562) with: {'C': 2.0, 'kernel': 'rbf'}\n",
        "\n",
        "0.733456 (0.090179) with: {'C': 2.0, 'kernel': 'sigmoid'}"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vasawFB8n6If"
      },
      "source": [
        "We can see the most accurate configuration was SVM with an RBF kernel and a C value\n",
        "of 2.0. The accuracy 87.2426% is seemingly better than what KNN could achieve."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RzEH65akoEsF"
      },
      "source": [
        "**Ensemble Methods**\n",
        "\n",
        "Another way that we can improve the performance of algorithms on this problem is by using\n",
        "ensemble methods. In this section we will evaluate four different ensemble machine learning\n",
        "algorithms, two boosting and two bagging methods:\n",
        "\n",
        "1. **Boosting Methods**: AdaBoost (AB) and Gradient Boosting (GBM).\n",
        "2. **Bagging Methods**: Random Forests (RF) and Extra Trees (ET).\n",
        "\n",
        "We will use the same test harness as before, 10-fold cross validation. No data standardization\n",
        "is used in this case because all four ensemble algorithms are based on decision trees that are\n",
        "less sensitive to data distributions.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "p251Ak7fnQJ1",
        "outputId": "8e47c5a9-516d-42b3-86e8-338d54c7df5c"
      },
      "source": [
        "# ensembles\n",
        "ensembles = []\n",
        "ensembles.append(('AB', AdaBoostClassifier()))\n",
        "ensembles.append(('GBM', GradientBoostingClassifier()))\n",
        "ensembles.append(('RF', RandomForestClassifier()))\n",
        "ensembles.append(('ET', ExtraTreesClassifier()))\n",
        "results = []\n",
        "names = []\n",
        "for name, model in ensembles:\n",
        "  kfold = KFold(n_splits=num_folds, random_state=seed)\n",
        "  cv_results = cross_val_score(model, X_train, Y_train, cv=kfold, scoring=scoring)\n",
        "  results.append(cv_results)\n",
        "  names.append(name)\n",
        "  msg = \"%s: %f (%f)\" % (name, cv_results.mean(), cv_results.std())\n",
        "  print(msg)"
      ],
      "execution_count": 60,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "AB: 0.813971 (0.066017)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "GBM: 0.854044 (0.106453)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "RF: 0.818015 (0.078684)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_split.py:296: FutureWarning: Setting a random_state has no effect since shuffle is False. This will raise an error in 0.24. You should leave random_state to its default (None), or set shuffle=True.\n",
            "  FutureWarning\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "ET: 0.847794 (0.093026)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FwG_O3MQoUe1"
      },
      "source": [
        "Running the example provides the following accuracy scores.\n",
        "\n",
        "AB: 0.819853 (0.058293)\n",
        "\n",
        "GBM: 0.853676 (0.100318)\n",
        "\n",
        "RF: 0.825368 (0.042087)\n",
        "\n",
        "ET: 0.847794 (0.093026)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "z4B-IgziouIh"
      },
      "source": [
        "We can see that both boosting techniques provide strong accuracy scores in the low 80s (%)\n",
        "with default configurations. We can plot the distribution of accuracy scores across the cross\n",
        "validation folds.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 294
        },
        "id": "XCgxReQvoRWp",
        "outputId": "409288ee-ab0b-4fa4-dad9-aa852cdb8976"
      },
      "source": [
        "# Compare Algorithms\n",
        "fig = pyplot.figure()\n",
        "fig.suptitle('Ensemble Algorithm Comparison')\n",
        "ax = fig.add_subplot(111)\n",
        "pyplot.boxplot(results)\n",
        "ax.set_xticklabels(names)\n",
        "pyplot.show()"
      ],
      "execution_count": 61,
      "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"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ldKq1DO0o4cQ"
      },
      "source": [
        "The results suggest GBM and ET may be worthy of further study, with a strong mean and a spread\n",
        "that skews up towards high 90s (%) in accuracy."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VyxDAbjDo-Xm"
      },
      "source": [
        "**Finalize Model**\n",
        "\n",
        "The SVM showed the most promise as a low complexity and stable model for this problem. In\n",
        "this section we will finalize the model by training it on the entire training dataset and make\n",
        "predictions for the hold-out validation dataset to confirm our findings. A part of the findings was\n",
        "that SVM performs better when the dataset is standardized so that all attributes have a mean\n",
        "value of zero and a standard deviation of one. We can calculate this from the entire training\n",
        "dataset and apply the same transform to the input attributes from the validation dataset.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HcaKf4Imowja",
        "outputId": "502afa1c-d79e-407b-f0a0-fa46ea3e88f4"
      },
      "source": [
        "# prepare the model\n",
        "scaler = StandardScaler().fit(X_train)\n",
        "rescaledX = scaler.transform(X_train)\n",
        "model = SVC(C=1.5)\n",
        "model.fit(rescaledX, Y_train)\n",
        "# estimate accuracy on validation dataset\n",
        "rescaledValidationX = scaler.transform(X_validation)\n",
        "predictions = model.predict(rescaledValidationX)\n",
        "print(accuracy_score(Y_validation, predictions))\n",
        "print(confusion_matrix(Y_validation, predictions))\n",
        "print(classification_report(Y_validation, predictions))"
      ],
      "execution_count": 62,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.8571428571428571\n",
            "[[23  4]\n",
            " [ 2 13]]\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           M       0.92      0.85      0.88        27\n",
            "           R       0.76      0.87      0.81        15\n",
            "\n",
            "    accuracy                           0.86        42\n",
            "   macro avg       0.84      0.86      0.85        42\n",
            "weighted avg       0.86      0.86      0.86        42\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YF1_2xt_pIns"
      },
      "source": [
        "We can see that we achieve an accuracy of nearly 86% on the held-out validation dataset. A\n",
        "score that matches closely to our expectations estimated above during the tuning of SVM."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PjxAvBiNpE3y",
        "outputId": "19b31f7c-fb11-46d6-e74b-c7c756175d73"
      },
      "source": [
        "scaler = StandardScaler().fit(X_train)\n",
        "rescaledX = scaler.transform(X_train)\n",
        "model = ExtraTreesClassifier(random_state=seed)\n",
        "model.fit(rescaledX, Y_train)\n",
        "# estimate accuracy on validation dataset\n",
        "rescaledValidationX = scaler.transform(X_validation)\n",
        "predictions = model.predict(rescaledValidationX)\n",
        "print(accuracy_score(Y_validation, predictions))\n",
        "print(confusion_matrix(Y_validation, predictions))\n",
        "print(classification_report(Y_validation, predictions))"
      ],
      "execution_count": 65,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9047619047619048\n",
            "[[25  2]\n",
            " [ 2 13]]\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           M       0.93      0.93      0.93        27\n",
            "           R       0.87      0.87      0.87        15\n",
            "\n",
            "    accuracy                           0.90        42\n",
            "   macro avg       0.90      0.90      0.90        42\n",
            "weighted avg       0.90      0.90      0.90        42\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zUfFyeL6pcD9"
      },
      "source": [
        ""
      ],
      "execution_count": 63,
      "outputs": []
    }
  ]
}