DMA/UE173085_Classifiers_Comparison.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.7"
    },
    "colab": {
      "name": "UE173094_Classifiers_Comparison.ipynb",
      "provenance": [],
      "collapsed_sections": []
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "prBZjLAYtY4m",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Run some setup code for this notebook.\n",
        "import random\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from sklearn import datasets"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CHdKRFYBtY4t",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#from sklearn.cross_validation import train_test_split\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "from sklearn.svm import SVC\n",
        "from sklearn.gaussian_process import GaussianProcessClassifier\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.neural_network import MLPClassifier\n",
        "from sklearn.ensemble import AdaBoostClassifier\n",
        "from sklearn.naive_bayes import GaussianNB\n",
        "from sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis\n",
        "from sklearn.gaussian_process.kernels import RBF\n",
        "from sklearn import metrics"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SLEekdc3tY4w",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "iris = datasets.load_iris()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y01lkF5CtY42",
        "colab_type": "code",
        "outputId": "c57d5060-105b-42fd-a6b8-793031c42303",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        }
      },
      "source": [
        "print(iris.keys())\n",
        "print(\"Header : \", iris.feature_names)\n",
        "print(\"Target : \", iris.target[np.random.choice(150,20)])\n",
        "print(iris.data.shape)"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "dict_keys(['data', 'target', 'target_names', 'DESCR', 'feature_names', 'filename'])\n",
            "Header :  ['sepal length (cm)', 'sepal width (cm)', 'petal length (cm)', 'petal width (cm)']\n",
            "Target :  [0 0 1 0 1 1 0 0 1 1 2 0 1 1 0 1 1 0 2 1]\n",
            "(150, 4)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GAbLUZxbIQkk",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "outputId": "97e3e0ab-5591-4247-c9e8-8e92273896b6"
      },
      "source": [
        "iris_df = pd.DataFrame(iris.data, columns=iris.feature_names)\n",
        "iris_df.head()"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>sepal length (cm)</th>\n",
              "      <th>sepal width (cm)</th>\n",
              "      <th>petal length (cm)</th>\n",
              "      <th>petal width (cm)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>5.1</td>\n",
              "      <td>3.5</td>\n",
              "      <td>1.4</td>\n",
              "      <td>0.2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>4.9</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.4</td>\n",
              "      <td>0.2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>4.7</td>\n",
              "      <td>3.2</td>\n",
              "      <td>1.3</td>\n",
              "      <td>0.2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4.6</td>\n",
              "      <td>3.1</td>\n",
              "      <td>1.5</td>\n",
              "      <td>0.2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5.0</td>\n",
              "      <td>3.6</td>\n",
              "      <td>1.4</td>\n",
              "      <td>0.2</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   sepal length (cm)  sepal width (cm)  petal length (cm)  petal width (cm)\n",
              "0                5.1               3.5                1.4               0.2\n",
              "1                4.9               3.0                1.4               0.2\n",
              "2                4.7               3.2                1.3               0.2\n",
              "3                4.6               3.1                1.5               0.2\n",
              "4                5.0               3.6                1.4               0.2"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AFU6zO3HIUiN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "029f25e7-daa7-4377-db10-a18d7d0b5745"
      },
      "source": [
        "# Split Training, Test Data\n",
        "\n",
        "num_train,_ = iris.data.shape\n",
        "\n",
        "indices = np.random.choice(num_train, num_train)\n",
        "\n",
        "train_X = iris.data[indices[:-20]].astype(\"float32\")\n",
        "train_y = iris.target[indices[:-20]]\n",
        "test_X = iris.data[indices[-20:]].astype(\"float32\")\n",
        "test_y = iris.target[indices[-20:]]\n",
        "\n",
        "print(train_X.shape)\n",
        "print(train_y.shape)\n",
        "print(test_X.shape)\n",
        "print(test_y.shape)"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(130, 4)\n",
            "(130,)\n",
            "(20, 4)\n",
            "(20,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9qrWRvY6tY5S",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "## Another way\n",
        "from sklearn.model_selection import train_test_split\n",
        "X = iris.data\n",
        "y = iris.target\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.4, random_state=4)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7GgPLS9s4sKD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "iris_accuracy = []"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lN6B-2sltY5i",
        "colab_type": "code",
        "outputId": "b81daccb-cfa5-45a2-ec52-099a9c0b7ddd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "knn = KNeighborsClassifier(n_neighbors=3)\n",
        "knn.fit(X_train, y_train)\n",
        "y_pred=knn.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "iris_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9666666666666667\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ixaHE9kKyuui",
        "colab_type": "code",
        "outputId": "5e223545-995c-4018-8590-ab0a392372b6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from sklearn.svm import SVC\n",
        "svc_clf = SVC(kernel=\"linear\", C=0.025)\n",
        "svc_clf.fit(X_train, y_train)\n",
        "y_pred = svc_clf.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "iris_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9666666666666667\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pWxMWaApyvvT",
        "colab_type": "code",
        "outputId": "e75dbfe9-178e-4f3e-c76b-94914b320fb0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from sklearn.svm import SVC\n",
        "svc_clf = SVC(gamma=2, C=1)\n",
        "svc_clf.fit(X_train, y_train)\n",
        "y_pred = svc_clf.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "iris_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9666666666666667\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2E0NulPky5bb",
        "colab_type": "code",
        "outputId": "277d3d08-1f5e-4870-aec5-5536d512ccda",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "kernel = 1.0 * RBF(1.0)\n",
        "gpc = GaussianProcessClassifier(kernel=kernel)\n",
        "gpc.fit(X_train, y_train)\n",
        "y_pred = gpc.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "iris_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9333333333333333\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BUxwq_sDz5-K",
        "colab_type": "code",
        "outputId": "be3a344a-5933-4b8a-a3a4-16253b02b23d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "dtc = DecisionTreeClassifier(max_depth=5)\n",
        "dtc.fit(X_train, y_train)\n",
        "y_pred = dtc.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "iris_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9666666666666667\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0INPigCf0cF5",
        "colab_type": "code",
        "outputId": "c2b950d2-1e74-4880-fa5d-9efcc98e927f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "rfc = RandomForestClassifier(max_depth=5, n_estimators=10, max_features=1)\n",
        "rfc.fit(X_train, y_train)\n",
        "y_pred = rfc.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "iris_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9333333333333333\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mUmaCKz312Ow",
        "colab_type": "code",
        "outputId": "29cede67-96a9-40de-f6b4-4176295812a9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 88
        }
      },
      "source": [
        "mpl = MLPClassifier(alpha=1)\n",
        "mpl.fit(X_train, y_train)\n",
        "y_pred = mpl.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "iris_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9333333333333333\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:571: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
            "  % self.max_iter, ConvergenceWarning)\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hsxtUi1I15B3",
        "colab_type": "code",
        "outputId": "a381ab83-bf46-4382-bdb9-14e15c2b1b52",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "ada = AdaBoostClassifier()\n",
        "ada.fit(X_train, y_train)\n",
        "y_pred = ada.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "iris_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9333333333333333\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SpAk_A9c16k4",
        "colab_type": "code",
        "outputId": "485c8da8-ffd5-4214-c100-7f79d5cc6649",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "gnb = GaussianNB()\n",
        "gnb.fit(X_train, y_train)\n",
        "y_pred = gnb.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "iris_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9666666666666667\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ri4yvxHq18K3",
        "colab_type": "code",
        "outputId": "b4a7c0e4-8fb0-47eb-a210-3214bdf0b4f7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "qda = QuadraticDiscriminantAnalysis()\n",
        "qda.fit(X_train, y_train)\n",
        "y_pred = qda.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "iris_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YELPV8lNtY5q",
        "colab_type": "code",
        "outputId": "f8d6eef7-0db0-436f-f4fe-a19f7f4ae850",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 120
        }
      },
      "source": [
        "############################################\n",
        "########### D I G I T S ####################\n",
        "############################################\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "digits = datasets.load_digits()\n",
        "_, axes = plt.subplots(1, 4)\n",
        "images_and_labels = list(zip(digits.images, digits.target))\n",
        "for ax, (image, label) in zip(axes[:], images_and_labels[:4]):\n",
        "    ax.set_axis_off()\n",
        "    ax.imshow(image, cmap=plt.cm.gray_r, interpolation='nearest')\n",
        "    ax.set_title('Training: %i' % label)\n",
        "\n",
        "n_samples = len(digits.images)\n",
        "data = digits.images.reshape((n_samples, -1))\n",
        "\n",
        "X_train, X_test, y_train, y_test = train_test_split(data, digits.target, test_size=0.2, shuffle=False)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9yjtjbCU5X7I",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "digits_accuracy = []"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "784b668f-bae8-4185-c3e2-393da1cc6f55",
        "id": "6rhYWtid2ZaE",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "knn = KNeighborsClassifier(n_neighbors=3)\n",
        "knn.fit(X_train, y_train)\n",
        "y_pred=knn.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "digits_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9666666666666667\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "4acf3572-af2a-428c-b660-ca237329e37d",
        "id": "0X22TKYy2Zah",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from sklearn.svm import SVC\n",
        "svc_clf = SVC(kernel=\"linear\", C=0.025)\n",
        "svc_clf.fit(X_train, y_train)\n",
        "y_pred = svc_clf.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "digits_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9305555555555556\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "19c1e39b-7200-497c-90b3-8cd9ab93424c",
        "id": "5CN4yP_w2Zal",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from sklearn.svm import SVC\n",
        "svc_clf = SVC(gamma=2, C=1)\n",
        "svc_clf.fit(X_train, y_train)\n",
        "y_pred = svc_clf.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "digits_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.18611111111111112\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "a5a4b595-7570-4459-8540-600647c8eb5a",
        "id": "NWY99-a02Zao",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "kernel = 1.0 * RBF(1.0)\n",
        "gpc = GaussianProcessClassifier(kernel=kernel)\n",
        "gpc.fit(X_train, y_train)\n",
        "y_pred = gpc.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "digits_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.10277777777777777\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "9facd096-dfc3-4c30-db01-416a2d69b619",
        "id": "WPAs5MCo2Zaq",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "dtc = DecisionTreeClassifier(max_depth=5)\n",
        "dtc.fit(X_train, y_train)\n",
        "y_pred = dtc.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "digits_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.6222222222222222\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "cb29fe93-7616-4570-a39b-9da9f33aac6c",
        "id": "u2CxmMtl2Zat",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "rfc = RandomForestClassifier(max_depth=5, n_estimators=10, max_features=1)\n",
        "rfc.fit(X_train, y_train)\n",
        "y_pred = rfc.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "digits_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.7861111111111111\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "44910151-1631-4303-8e8b-70133dcb7a4b",
        "id": "J5hXlsmt2Zav",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 88
        }
      },
      "source": [
        "mpl = MLPClassifier(alpha=1)\n",
        "mpl.fit(X_train, y_train)\n",
        "y_pred = mpl.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "digits_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9083333333333333\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:571: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
            "  % self.max_iter, ConvergenceWarning)\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "3c4ca3e5-9ea7-4e0a-c2d3-ab3bee0caef0",
        "id": "9xih53-Y2Zay",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "ada = AdaBoostClassifier()\n",
        "ada.fit(X_train, y_train)\n",
        "y_pred = ada.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "digits_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.2777777777777778\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "5f71f58c-3e55-46d2-a0c9-8cd4e3b819c0",
        "id": "S3f8gn9Q2Za1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "gnb = GaussianNB()\n",
        "gnb.fit(X_train, y_train)\n",
        "y_pred = gnb.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "digits_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.8138888888888889\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "7f2c07c3-1c77-460a-fd52-de75ce32901e",
        "id": "b741jHFs2Za3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "qda = QuadraticDiscriminantAnalysis()\n",
        "qda.fit(X_train, y_train)\n",
        "y_pred = qda.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "digits_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.7861111111111111\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/sklearn/discriminant_analysis.py:691: UserWarning: Variables are collinear\n",
            "  warnings.warn(\"Variables are collinear\")\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iypZrt7w7YR-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "############################################\n",
        "########### B O S T O N ####################\n",
        "############################################\n",
        "boston = datasets.load_boston()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3AIt9hOS6JNw",
        "colab_type": "code",
        "outputId": "6acf0a5f-733c-4fbc-f2d4-a60ba1a9cbe8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        }
      },
      "source": [
        "data = boston.data\n",
        "target = boston.target\n",
        "df = pd.DataFrame(target)\n",
        "df.describe()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>0</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>506.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>22.532806</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>9.197104</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>5.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>17.025000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>21.200000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>25.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>50.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                0\n",
              "count  506.000000\n",
              "mean    22.532806\n",
              "std      9.197104\n",
              "min      5.000000\n",
              "25%     17.025000\n",
              "50%     21.200000\n",
              "75%     25.000000\n",
              "max     50.000000"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 270
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UdbujJCa7Vbl",
        "colab_type": "code",
        "outputId": "10151b60-b49e-4b42-e8d4-c7acb1f9cc53",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "low = 0\n",
        "medium = 1\n",
        "high = 2\n",
        "new_target = []\n",
        "for x in target:\n",
        "  if x<20:\n",
        "    new_target.append(low)\n",
        "  elif x<35:\n",
        "    new_target.append(medium)\n",
        "  else:\n",
        "    new_target.append(high)   \n",
        "print(new_target[:10])\n",
        "print(target[:10])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[1, 1, 1, 1, 2, 1, 1, 1, 0, 0]\n",
            "[24.  21.6 34.7 33.4 36.2 28.7 22.9 27.1 16.5 18.9]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4wIC3Kxu7e1a",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X_train, X_test, y_train, y_test = train_test_split(data, new_target, test_size=0.2, shuffle=False)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "SNU6nzX48ouO",
        "colab": {}
      },
      "source": [
        "boston_accuracy = []"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "4c223758-b2bf-4eed-c8c6-d1662701bc56",
        "id": "qr901h9C8ouj",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "knn = KNeighborsClassifier(n_neighbors=3)\n",
        "knn.fit(X_train, y_train)\n",
        "y_pred=knn.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "boston_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.7352941176470589\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "b38794ee-569d-49e9-d59b-7647831a87fb",
        "id": "AT0mWEwW8ou0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from sklearn.svm import SVC\n",
        "svc_clf = SVC(kernel=\"linear\", C=0.025)\n",
        "svc_clf.fit(X_train, y_train)\n",
        "y_pred = svc_clf.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "boston_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.8333333333333334\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "bdd52dc5-2e72-4fb6-b3db-f83ec6a6158b",
        "id": "ZBA6BeYe8ou6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from sklearn.svm import SVC\n",
        "svc_clf = SVC(gamma=2, C=1)\n",
        "svc_clf.fit(X_train, y_train)\n",
        "y_pred = svc_clf.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "boston_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.2549019607843137\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "dfe9d63d-1e15-4a99-ecc0-effe500fa8c1",
        "id": "bBgQDTxy8ou-",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "kernel = 1.0 * RBF(1.0)\n",
        "gpc = GaussianProcessClassifier(kernel=kernel)\n",
        "gpc.fit(X_train, y_train)\n",
        "y_pred = gpc.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "boston_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.6568627450980392\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "d1e3f578-432a-4d5b-e1ae-8628f940086c",
        "id": "uNvgUI9W8ovC",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "dtc = DecisionTreeClassifier(max_depth=5)\n",
        "dtc.fit(X_train, y_train)\n",
        "y_pred = dtc.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "boston_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.8725490196078431\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "a54c2c43-471f-41ab-dd8c-010e6f1b8c7e",
        "id": "vxSboLi58ovG",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "rfc = RandomForestClassifier(max_depth=5, n_estimators=10, max_features=1)\n",
        "rfc.fit(X_train, y_train)\n",
        "y_pred = rfc.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "boston_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.8137254901960784\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "d403f1dc-9d5d-4788-d82c-ec29c211b8e6",
        "id": "YnQXLcRZ8ovJ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "mpl = MLPClassifier(alpha=1)\n",
        "mpl.fit(X_train, y_train)\n",
        "y_pred = mpl.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "boston_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.6862745098039216\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "22b905bb-04f2-4417-a165-13b58387ccec",
        "id": "bNdjsvd_8ovN",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "ada = AdaBoostClassifier()\n",
        "ada.fit(X_train, y_train)\n",
        "y_pred = ada.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "boston_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.22549019607843138\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "1e4fa5ee-6b30-40f1-d1ad-22eed34ef798",
        "id": "2iHUP3pp8ovP",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "gnb = GaussianNB()\n",
        "gnb.fit(X_train, y_train)\n",
        "y_pred = gnb.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "boston_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.7843137254901961\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "c8c31d4b-3c32-4ad4-ed15-1cebb869e066",
        "id": "0trGmOmR8ovR",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "qda = QuadraticDiscriminantAnalysis()\n",
        "qda.fit(X_train, y_train)\n",
        "y_pred = qda.predict(X_test)\n",
        "score = metrics.accuracy_score(y_test, y_pred)\n",
        "boston_accuracy.append(score)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.7941176470588235\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dcXnO-8G4eL4",
        "colab_type": "code",
        "outputId": "4adc561f-0f9e-4733-c7ed-bd83583dd3e0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 662
        }
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "classifiers = ['KNeighborsClassifier(3)', 'SVC(kernel=\"linear\", C=0.025)', 'SVC(gamma=2, C=1)', 'GaussianProcessClassifier(1.0 * RBF(1.0))', 'DecisionTreeClassifier(max_depth=5)', 'RandomForestClassifier(max_depth=5, n_estimators=10, max_features=1)', 'MLPClassifier(alpha=1)', 'AdaBoostClassifier()', 'GaussianNB()', 'QuadraticDiscriminantAnalysis()']\n",
        "plt.plot(classifiers, iris_accuracy, color='g')\n",
        "plt.plot(classifiers, digits_accuracy, color='orange')\n",
        "plt.plot(classifiers, boston_accuracy, color='blue')\n",
        "plt.xticks(rotation=90)\n",
        "plt.xlabel('Classifier')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.title('Classifier accuracy comparison')\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "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"
          }
        }
      ]
    }
  ]
}