kemalseker_HW3_Logistic_Regression_toSolve.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "kemalseker_HW3-Logistic Regression-toSolve.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "4ZzBwrMglgYl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "from sklearn.naive_bayes import GaussianNB\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import classification_report, confusion_matrix, accuracy_score\n",
        "import numpy as np"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dxKFHk_HynHe",
        "colab_type": "text"
      },
      "source": [
        "# [25 pts] A Toy Example: Decision Boundary and Conditional Independence Assumption"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fp4nIW7QssBo",
        "colab_type": "text"
      },
      "source": [
        "### Gaussian Distributed Data fits better to Gaussian Naive Bayesian rather than Logistic Regression, unfortunately that is not the case most of the time.\n",
        "Now, imagine we have two artificial dataset. Both are drawn from Gaussian distribution. One of the dataset is with standard deviation 1 and the other is 5. Each cluster is conditionally independent from each other.\n",
        "\n",
        "make_blobs function samples data points from gaussian distribution."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "tPgLMm3csrD6",
        "colab": {}
      },
      "source": [
        "from sklearn.datasets import make_blobs\n",
        "data1, label1 = make_blobs(n_samples=1500 , cluster_std=1 , centers=2 , random_state=1, n_features=2) # code comes here\n",
        "data2, label2 = make_blobs(n_samples=1500 , cluster_std=5 , centers=2, random_state=1, n_features=2)# code comes here"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VrJE4Mxttpqo",
        "colab_type": "text"
      },
      "source": [
        "Let's split the datasets into train and test."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "Yq25ZZR4sqrC",
        "colab": {}
      },
      "source": [
        "# code comes here\n",
        "x_train, x_val, y_train, y_val = train_test_split(data1, label1, test_size=0.2)\n",
        "# code comes here\n",
        "x2_train, x2_val, y2_train, y2_val = train_test_split(data2, label2, test_size=0.2)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Buvhwfz8tub7",
        "colab_type": "text"
      },
      "source": [
        "Plot the first dataset with standard deviation 1."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "F8Q_5tensqas",
        "outputId": "8bb95ebe-1ce8-4896-d155-e939e9fc891c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "plt.scatter(data1[:,0], data1[:,1])\n",
        "plt.title('Scatter plot data with standard deviation=1')\n",
        "plt.xlabel('x')\n",
        "plt.ylabel('y')\n",
        "plt.show()"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEWCAYAAABv+EDhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "B3OgT3DVt0JO",
        "colab_type": "text"
      },
      "source": [
        "Plot the first dataset with standard deviation 4."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "kMYmZLvfsqJK",
        "outputId": "8227e256-e9f8-41da-8fb3-dc9b1739a11a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "plt.scatter(data2[:,0], data2[:,1])\n",
        "plt.title('Scatter plot data with standard deviation=5')\n",
        "plt.xlabel('x')\n",
        "plt.ylabel('y')\n",
        "plt.show()"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAEWCAYAAACNJFuYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\noYb/H907AHUUPG4TAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Al6vGgvUtZf2",
        "colab_type": "text"
      },
      "source": [
        "Train a Gaussian Naive Bayesian and Logistic Regression with the 1st dataset."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "vfsu5AaIsp1Y",
        "outputId": "12409a60-b846-44d8-d7ef-f526e75ae7e7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# code comes here\n",
        "gnb = GaussianNB();\n",
        "# code comes here\n",
        "gnb.fit(x_train , y_train)"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "GaussianNB(priors=None, var_smoothing=1e-09)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "adUJfl64spmM",
        "outputId": "2a4723ad-3328-45a4-dc3a-685c55a23ad0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 101
        }
      },
      "source": [
        "# code comes here\n",
        "logreg = LogisticRegression()\n",
        "# code comes here\n",
        "logreg.fit(x_train,y_train)"
      ],
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
              "                   intercept_scaling=1, l1_ratio=None, max_iter=100,\n",
              "                   multi_class='auto', n_jobs=None, penalty='l2',\n",
              "                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,\n",
              "                   warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Iw3NUUHfuSf4",
        "colab_type": "text"
      },
      "source": [
        "### Perfect Decision Boundary"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "lCOG0Ii0spWm",
        "outputId": "aedf84e9-395b-4b46-f3da-32b9761ded6c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 185
        }
      },
      "source": [
        "# Predict\n",
        "pred = gnb.predict(x_val)\n",
        "print(\"Classification Report for Naive Bayesian:\")\n",
        "print(classification_report(y_val, pred))"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Classification Report for Naive Bayesian:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       1.00      1.00      1.00       154\n",
            "           1       1.00      1.00      1.00       146\n",
            "\n",
            "    accuracy                           1.00       300\n",
            "   macro avg       1.00      1.00      1.00       300\n",
            "weighted avg       1.00      1.00      1.00       300\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "hV1gbmlfspHg",
        "outputId": "3e03b7fd-ce29-410f-9d73-c2f3192daf5a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 185
        }
      },
      "source": [
        "# Predict\n",
        "pred2 = logreg.predict(x_val)\n",
        "print(\"Classification Report for Logistic Regression:\")\n",
        "print(classification_report(y_val , pred2))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Classification Report for Logistic Regression:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       1.00      1.00      1.00       144\n",
            "           1       1.00      1.00      1.00       156\n",
            "\n",
            "    accuracy                           1.00       300\n",
            "   macro avg       1.00      1.00      1.00       300\n",
            "weighted avg       1.00      1.00      1.00       300\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "G6_Ss-WXt_f6",
        "colab_type": "text"
      },
      "source": [
        "### Both algorithm perfectly separate two data clusters for 1st dataset with standard deviation 1. The data points are linearly separable."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "w-BrV7Ftso30",
        "outputId": "5015b166-dfc4-491e-e989-f8814a8ae4d6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# code comes here\n",
        "gnb2 = GaussianNB();\n",
        "# code comes here\n",
        "gnb2.fit(x2_train , y2_train)"
      ],
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "GaussianNB(priors=None, var_smoothing=1e-09)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "s9TGNXqOsooC",
        "outputId": "550ec727-6282-4cc5-c355-c7774de01212",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 101
        }
      },
      "source": [
        "# code comes here\n",
        "logreg2 = LogisticRegression()\n",
        "# code comes here\n",
        "logreg2.fit(x2_train,y2_train)"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
              "                   intercept_scaling=1, l1_ratio=None, max_iter=100,\n",
              "                   multi_class='auto', n_jobs=None, penalty='l2',\n",
              "                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,\n",
              "                   warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "ezSLARzIsoY0",
        "outputId": "bf1d3d94-182e-45c6-d460-bf31668cc977",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 185
        }
      },
      "source": [
        "# Predict\n",
        "pred3 = gnb2.predict(x2_val)\n",
        "print(\"Classification Report for Naive Bayesian:\")\n",
        "print(classification_report(y2_val, pred3))"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Classification Report for Naive Bayesian:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.91      0.89      0.90       167\n",
            "           1       0.87      0.89      0.88       133\n",
            "\n",
            "    accuracy                           0.89       300\n",
            "   macro avg       0.89      0.89      0.89       300\n",
            "weighted avg       0.89      0.89      0.89       300\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "ZzZ3KzaosoDM",
        "outputId": "3ff7ee3f-eb17-416e-dbab-4bfd25b3bda2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 185
        }
      },
      "source": [
        "# Predict\n",
        "pred4 = logreg.predict(x2_val)\n",
        "print(\"Classification Report for Logistic Regression:\")\n",
        "print(classification_report(y2_val, pred4))"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Classification Report for Logistic Regression:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.91      0.90      0.91       167\n",
            "           1       0.88      0.89      0.88       133\n",
            "\n",
            "    accuracy                           0.90       300\n",
            "   macro avg       0.90      0.90      0.90       300\n",
            "weighted avg       0.90      0.90      0.90       300\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MikCmwPI_JU7",
        "colab_type": "code",
        "outputId": "eb30cc4c-d404-4776-b482-bb6214348a2f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 70
        }
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "from matplotlib.colors import ListedColormap\n",
        "import seaborn as sns\n",
        "\n",
        "%matplotlib inline\n",
        "\n",
        "# Import statements required for Plotly \n",
        "import plotly.offline as py\n",
        "py.init_notebook_mode(connected=True)\n",
        "import plotly.graph_objs as go\n",
        "from plotly import tools"
      ],
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
            "  import pandas.util.testing as tm\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "        <script type=\"text/javascript\">\n",
              "        window.PlotlyConfig = {MathJaxConfig: 'local'};\n",
              "        if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}\n",
              "        if (typeof require !== 'undefined') {\n",
              "        require.undef(\"plotly\");\n",
              "        requirejs.config({\n",
              "            paths: {\n",
              "                'plotly': ['https://cdn.plot.ly/plotly-latest.min']\n",
              "            }\n",
              "        });\n",
              "        require(['plotly'], function(Plotly) {\n",
              "            window._Plotly = Plotly;\n",
              "        });\n",
              "        }\n",
              "        </script>\n",
              "        "
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MrawrQVUi0bE",
        "colab_type": "code",
        "outputId": "54a5d31d-e5cb-442f-92ec-9f5104674cf2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 334
        }
      },
      "source": [
        "X_set, y_set = x_train, y_train\n",
        "X1, X2 = np.meshgrid(np.arange(start = X_set[:, 0].min() - 1, stop = X_set[:, 0].max() + 1, step = 0.1),\n",
        "                     np.arange(start = X_set[:, 1].min() - 1, stop = X_set[:, 1].max() + 1, step = 0.1))\n",
        "plt.contourf(X1, X2, gnb.predict(np.array([X1.ravel(), X2.ravel()]).T).reshape(X1.shape),\n",
        "             alpha = 0.75, cmap = ListedColormap(('yellow', 'white')))\n",
        "plt.xlim(X1.min(), X1.max())\n",
        "plt.ylim(X2.min(), X2.max())\n",
        "for i, j in enumerate(np.unique(y_set)):\n",
        "    plt.scatter(X_set[y_set == j, 0], X_set[y_set == j, 1],\n",
        "                c = ListedColormap(('red', 'green'))(i), label = j,marker='.')\n",
        "plt.title('Naive Bayes Classification our implementation(Dataset with Standart Deviation=1)')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "'c' argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with 'x' & 'y'.  Please use a 2-D array with a single row if you really want to specify the same RGB or RGBA value for all points.\n",
            "'c' argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with 'x' & 'y'.  Please use a 2-D array with a single row if you really want to specify the same RGB or RGBA value for all points.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAegAAAEICAYAAACUFGeOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "64wIzfW3l7nw",
        "colab_type": "code",
        "outputId": "83f2c98b-ca6a-42af-a699-9eb8edfac042",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 334
        }
      },
      "source": [
        "X_set, y_set = x2_train, y2_train\n",
        "X1, X2 = np.meshgrid(np.arange(start = X_set[:, 0].min() - 1, stop = X_set[:, 0].max() + 1, step = 0.1),\n",
        "                     np.arange(start = X_set[:, 1].min() - 1, stop = X_set[:, 1].max() + 1, step = 0.1))\n",
        "plt.contourf(X1, X2, gnb2.predict(np.array([X1.ravel(), X2.ravel()]).T).reshape(X1.shape),\n",
        "             alpha = 0.75, cmap = ListedColormap(('yellow', 'white')))\n",
        "plt.xlim(X1.min(), X1.max())\n",
        "plt.ylim(X2.min(), X2.max())\n",
        "for i, j in enumerate(np.unique(y_set)):\n",
        "    plt.scatter(X_set[y_set == j, 0], X_set[y_set == j, 1],\n",
        "                c = ListedColormap(('red', 'green'))(i), label = j,marker='.')\n",
        "plt.title('Naive Bayes Classification our implementation(Dataset with Standart Deviation=5)')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "'c' argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with 'x' & 'y'.  Please use a 2-D array with a single row if you really want to specify the same RGB or RGBA value for all points.\n",
            "'c' argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with 'x' & 'y'.  Please use a 2-D array with a single row if you really want to specify the same RGB or RGBA value for all points.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgUAAAEICAYAAADP6odOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EOEm_p0RukGE",
        "colab_type": "text"
      },
      "source": [
        "### Use the scatter plot and draw the perfect decision boundary on two scatter plot. Discuss what is linear separability, decision boundary, which datapoints are harder to separate. Discuss the accuries and the why which model performs better.\n",
        " \n",
        "\n",
        "### Please also read: [Equivalence of GNB and LR](https://appliedmachinelearning.blog/2019/09/30/equivalence-of-gaussian-naive-bayes-and-logistic-regression-an-explanation/)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tiF0MX4d5eYq",
        "colab_type": "text"
      },
      "source": [
        "**Discussion**\n",
        "\n",
        "Linear separability is where two data features can separeted by an linear line. Decision boundry is seperation of one region from other region or regions. First data with standart deviation one can more easily separated. Because GNB is more suitable for linear separable datas. But second data is harder to separate with GNB because it is not a linearly seperable. GNB assumes features are statistically independent. It models the actual distribution of each class. It is using joint probability distribution. But, logistic regression use conditional probability distribution. However, datasets we use are conditionally independent. So GNB can give us better results due to that fact. As we can see from classification reports GNB has better ‘recall’ for 1 and ‘f1-score’ for 0.     "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "r3NlcMcuybLG",
        "colab_type": "text"
      },
      "source": [
        "# [75pts] Logistic Regression and Naive Bayesian Comparison"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pTCdpwctybP_",
        "colab_type": "text"
      },
      "source": [
        "### The dataset\n",
        "We will use Kaggle dataset. This dataset contains around 200k news headlines from the year 2012 to 2018 obtained from HuffPost.\n",
        "\n",
        "You can [download.](https://www.kaggle.com/rmisra/news-category-dataset)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4XI0ssM-vESc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.feature_extraction.text import CountVectorizer,TfidfVectorizer"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tGxi2HdLRaSQ",
        "colab_type": "code",
        "outputId": "2acbbeac-0266-4a7a-b101-b09242c0a4a7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 121
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "O_qxdl1W0vwU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df=pd.read_json('/content/drive/My Drive/News_Category_Dataset_v2.json' , lines=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3vw87GC9HyCb",
        "colab_type": "text"
      },
      "source": [
        "## Select 4 categories: Politics, Wellness, Entertainment, Travel"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XC9FiyORcb7_",
        "colab_type": "text"
      },
      "source": [
        "use only 50K of data row"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EsGFclVkMsu7",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df = df.sample(50000)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IITl1l8uFS7-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "new_df = df[(df['category']== 'POLITICS') | (df['category']== 'WELLNESS') | (df['category']== 'ENTERTAINMENT') | (df['category']== 'TRAVEL')]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XDfkCxBnE6ll",
        "colab_type": "code",
        "outputId": "40b218d7-b57b-4d3e-c673-3f2b431a206d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 101
        }
      },
      "source": [
        "new_df['category'].value_counts()"
      ],
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "POLITICS         8150\n",
              "WELLNESS         4449\n",
              "ENTERTAINMENT    4014\n",
              "TRAVEL           2452\n",
              "Name: category, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JMNVxLYmJx9c",
        "colab_type": "text"
      },
      "source": [
        "Convert category names to digit labelling"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yDYfDUe4E6qY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y = (new_df['category'].to_numpy() == \"WELLNESS\")*1 + (new_df['category'].to_numpy() == \"ENTERTAINMENT\")*2 + (new_df['category'].to_numpy() == \"TRAVEL\")*3"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kbzCIMKqKD-S",
        "colab_type": "text"
      },
      "source": [
        "Merge headlines with short descriptions"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I02ZKccjE6tT",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X = new_df['short_description'] + ' '+ new_df['headline']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9CI9QnFpKY-z",
        "colab_type": "text"
      },
      "source": [
        "### Create Tf-Idf model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FvsbWnHgE6wK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.feature_extraction.text import CountVectorizer\n",
        "\n",
        "# code comes here\n",
        "count_vect = CountVectorizer(lowercase=True, stop_words='english')\n",
        "# code comes here\n",
        "X_train_counts = count_vect.fit_transform(X)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "egpmnHfOE6za",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.feature_extraction.text import TfidfTransformer\n",
        "\n",
        "# code comes here\n",
        "tfidf_transformer = TfidfTransformer()\n",
        "# code comes here\n",
        "X_train_tf = tfidf_transformer.fit_transform(X_train_counts)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CvqXQFBoKcau",
        "colab_type": "text"
      },
      "source": [
        "Split train and test data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1fAe2Ao8E63a",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "# code comes here\n",
        "x_train, x_val, y_train, y_val = train_test_split(X_train_tf, y, test_size=0.2 )"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mmGiX6u6Lf2a",
        "colab_type": "text"
      },
      "source": [
        "### Gaussian Naive Bayesian"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ry0v_z27Ku2c",
        "colab_type": "code",
        "outputId": "f2f7d980-114a-4f79-ab5e-3e0589e988dd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# code comes here\n",
        "GNB = GaussianNB()\n",
        "# code comes here\n",
        "GNB.fit(x_train.toarray(), y_train)"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "GaussianNB(priors=None, var_smoothing=1e-09)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y9OqGk-9Ku7Z",
        "colab_type": "code",
        "outputId": "2964c3bf-b4e4-4241-d151-1976564adeab",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 218
        }
      },
      "source": [
        "# Predict\n",
        "print(\"Classification Report for Naive Bayesian:\")\n",
        "print(classification_report(GNB.predict(x_val.toarray()), y_val))"
      ],
      "execution_count": 50,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Classification Report for Naive Bayesian:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.75      0.79      0.77      1585\n",
            "           1       0.68      0.70      0.69       851\n",
            "           2       0.65      0.62      0.63       810\n",
            "           3       0.66      0.56      0.61       567\n",
            "\n",
            "    accuracy                           0.70      3813\n",
            "   macro avg       0.68      0.67      0.68      3813\n",
            "weighted avg       0.70      0.70      0.70      3813\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hPc237T_AbPO",
        "colab_type": "code",
        "outputId": "b3b4abea-5d4c-4e6a-aefa-e5b7b550c267",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 166
        }
      },
      "source": [
        "labels = np.unique(y)\n",
        "resultsGNB = confusion_matrix(y_val,GNB.predict(x_val.toarray()) ,labels=labels ) \n",
        "pd.DataFrame(resultsGNB, index=labels, columns=labels)"
      ],
      "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",
              "      <th>1</th>\n",
              "      <th>2</th>\n",
              "      <th>3</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1259</td>\n",
              "      <td>131</td>\n",
              "      <td>216</td>\n",
              "      <td>69</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>137</td>\n",
              "      <td>584</td>\n",
              "      <td>57</td>\n",
              "      <td>76</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>136</td>\n",
              "      <td>53</td>\n",
              "      <td>563</td>\n",
              "      <td>61</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>99</td>\n",
              "      <td>57</td>\n",
              "      <td>33</td>\n",
              "      <td>289</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      0    1    2    3\n",
              "0  1259  131  216   69\n",
              "1   137  584   57   76\n",
              "2   136   53  563   61\n",
              "3    99   57   33  289"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 176
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KMbfLFGqLxDd",
        "colab_type": "text"
      },
      "source": [
        "### 6) Logistic Regression"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pIBqMFP2Ku-5",
        "colab_type": "code",
        "outputId": "de585f7b-823b-46e8-b369-61ef84e915af",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 101
        }
      },
      "source": [
        "#code comes here\n",
        "LOGREG = LogisticRegression(max_iter=100000000)\n",
        "# code comes here\n",
        "LOGREG.fit(x_train.toarray(),y_train)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
              "                   intercept_scaling=1, l1_ratio=None, max_iter=100000000,\n",
              "                   multi_class='auto', n_jobs=None, penalty='l2',\n",
              "                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,\n",
              "                   warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 177
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NKxcvdTHKu5a",
        "colab_type": "code",
        "outputId": "f134da58-57f4-4b63-fbad-703cbb290782",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 230
        }
      },
      "source": [
        "# Predict\n",
        "print(\"Classification Report for Logistic Regression:\")\n",
        "print(classification_report(LOGREG.predict(x_val.toarray()) , y_val))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Classification Report for Logistic Regression:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.96      0.90      0.93      1726\n",
            "           1       0.91      0.90      0.90       917\n",
            "           2       0.85      0.90      0.87       752\n",
            "           3       0.78      0.92      0.85       432\n",
            "\n",
            "    accuracy                           0.90      3827\n",
            "   macro avg       0.88      0.90      0.89      3827\n",
            "weighted avg       0.91      0.90      0.90      3827\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Na5bzlqRBgpT",
        "colab_type": "code",
        "outputId": "62f49384-ad82-4c9b-fe63-818506ef7df7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 166
        }
      },
      "source": [
        "labels = np.unique(y)\n",
        "resultsLGR = confusion_matrix(y_val,LOGREG.predict(x_val.toarray()) , labels=labels ) \n",
        "#print (\"Confusion Matrix :\")\n",
        "#print(resultsLGR) \n",
        "pd.DataFrame(resultsLGR, index=labels, columns=labels)"
      ],
      "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",
              "      <th>1</th>\n",
              "      <th>2</th>\n",
              "      <th>3</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1604</td>\n",
              "      <td>26</td>\n",
              "      <td>34</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>42</td>\n",
              "      <td>759</td>\n",
              "      <td>31</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>79</td>\n",
              "      <td>22</td>\n",
              "      <td>675</td>\n",
              "      <td>8</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>31</td>\n",
              "      <td>48</td>\n",
              "      <td>42</td>\n",
              "      <td>402</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      0    1    2    3\n",
              "0  1604   26   34    6\n",
              "1    42  759   31   11\n",
              "2    79   22  675    8\n",
              "3    31   48   42  402"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 164
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "V70hLs_YNZXI",
        "colab_type": "text"
      },
      "source": [
        "### Observe Logistic Regression is much slower but more accurate. Discuss."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "695oyfFnv-T5",
        "colab_type": "text"
      },
      "source": [
        "**Report**\n",
        "\n",
        " We compare two machine learning models with each other. They are Logistic Regression and Gaussian Naive Bayes. We used an news paper data set with size 50K. We have 4 topics for the news and we try to estimate them. We have obtianed the best results with using Logistic Regression, giving a accuracy of %90 on test data. GNB has given more poor results if compared. It is giving a accuracy of %70. We have better result with Logistic Regression because our data has many correlated features. And logistic regression works better when there is more correlated features because it is using conditional probability distribution. \n",
        " In terms of speed GNB is faster than Logistic Regression. \n",
        "In logistic regression we have very high accuracy. We can see observe it from conditional matrix as well. For example, logistic regression estimate 1604 Politic articles true otherwise GNB's score is 1259.   \n",
        "\n"
      ]
    }
  ]
}