document_classification_stacking_voting.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Assignment_10_201911003_IR.ipynb",
      "provenance": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6on0swl4pqn-",
        "colab_type": "text"
      },
      "source": [
        "# **DATA PREPARATION**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "i6qn7ivFEJWM",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.datasets import load_files\n",
        "data = load_files('/content/bbc',shuffle=False,encoding='latin')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GS-3pPuoFcxf",
        "colab_type": "code",
        "outputId": "6d2c178e-5df5-4998-c5e7-e73238615814",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 194
        }
      },
      "source": [
        "import pandas as pd\n",
        "bbc_news_data = list(zip(data.data,data.target))\n",
        "bbc_news_dataframe = pd.DataFrame(data = bbc_news_data, columns=['TEXT','LABEL'])\n",
        "bbc_news_dataframe.head()"
      ],
      "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>TEXT</th>\n",
              "      <th>LABEL</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Ad sales boost Time Warner profit\\n\\nQuarterly...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Dollar gains on Greenspan speech\\n\\nThe dollar...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Yukos unit buyer faces loan claim\\n\\nThe owner...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>High fuel prices hit BA's profits\\n\\nBritish A...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Pernod takeover talk lifts Domecq\\n\\nShares in...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                TEXT  LABEL\n",
              "0  Ad sales boost Time Warner profit\\n\\nQuarterly...      0\n",
              "1  Dollar gains on Greenspan speech\\n\\nThe dollar...      0\n",
              "2  Yukos unit buyer faces loan claim\\n\\nThe owner...      0\n",
              "3  High fuel prices hit BA's profits\\n\\nBritish A...      0\n",
              "4  Pernod takeover talk lifts Domecq\\n\\nShares in...      0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yUUDHtOslNw9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "bbc_news_dataframe.to_csv('/content/drive/My Drive/Information_Retrieval/bbc_news_dataframe.csv')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Gu0FuOIfFRQq",
        "colab_type": "code",
        "outputId": "57c8f1e4-d77c-4607-a2b1-724f4fbcae8e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 194
        }
      },
      "source": [
        "import pandas as pd\n",
        "\n",
        "bbc_news_dataframe = pd.read_csv('/content/drive/My Drive/Information_Retrieval/bbc_news_dataframe.csv',index_col=False)\n",
        "bbc_news_dataframe.head()"
      ],
      "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>Unnamed: 0</th>\n",
              "      <th>TEXT</th>\n",
              "      <th>LABEL</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>Ad sales boost Time Warner profit\\n\\nQuarterly...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>Dollar gains on Greenspan speech\\n\\nThe dollar...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>Yukos unit buyer faces loan claim\\n\\nThe owner...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>High fuel prices hit BA's profits\\n\\nBritish A...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>Pernod takeover talk lifts Domecq\\n\\nShares in...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Unnamed: 0                                               TEXT  LABEL\n",
              "0           0  Ad sales boost Time Warner profit\\n\\nQuarterly...      0\n",
              "1           1  Dollar gains on Greenspan speech\\n\\nThe dollar...      0\n",
              "2           2  Yukos unit buyer faces loan claim\\n\\nThe owner...      0\n",
              "3           3  High fuel prices hit BA's profits\\n\\nBritish A...      0\n",
              "4           4  Pernod takeover talk lifts Domecq\\n\\nShares in...      0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Btm1ijnAIYk-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "bbc_news_train, bbc_news_test, bbc_news_train_labels, bbc_news_test_labels = train_test_split(bbc_news_dataframe['TEXT'], bbc_news_dataframe['LABEL'],\n",
        "                                                                                              stratify=bbc_news_dataframe['LABEL'], test_size=0.3,random_state = 10)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WHz1sAeQi1xJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "\n",
        "tfIdfVectorier = TfidfVectorizer(stop_words = 'english', min_df = 0)\n",
        "bbc_news_train_vectors = tfIdfVectorier.fit_transform(bbc_news_train)\n",
        "bbc_news_test_vectors = tfIdfVectorier.transform(bbc_news_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L6knEsfQkSWJ",
        "colab_type": "code",
        "outputId": "c5984338-a359-49c1-b904-b7de4efe2a1e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "bbc_news_train_vectors.shape,bbc_news_test_vectors.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((1557, 25108), (668, 25108))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NOWMjwP2F19G",
        "colab_type": "text"
      },
      "source": [
        "# **VOTING CLASSIFIER**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I-gba6jJZ3Ym",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "from sklearn.ensemble import RandomForestClassifier, VotingClassifier\n",
        "\n",
        "clf1 = LogisticRegression(C = 10, random_state=1)\n",
        "clf2 = RandomForestClassifier(n_estimators=100,random_state=1)\n",
        "clf3 = MultinomialNB(alpha = 0.1)\n",
        "\n",
        "eclf1 = VotingClassifier(estimators=[('lr', clf1), ('rf', clf2), ('mnb', clf3)], voting='hard')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rTlt2cNDcW81",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 568
        },
        "outputId": "5bbecc9b-4987-423e-a9da-0fcf3939ab69"
      },
      "source": [
        "eclf1.fit(bbc_news_train_vectors, bbc_news_train_labels)"
      ],
      "execution_count": 58,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "VotingClassifier(estimators=[('lr',\n",
              "                              LogisticRegression(C=10, class_weight=None,\n",
              "                                                 dual=False, fit_intercept=True,\n",
              "                                                 intercept_scaling=1,\n",
              "                                                 l1_ratio=None, max_iter=100,\n",
              "                                                 multi_class='auto',\n",
              "                                                 n_jobs=None, penalty='l2',\n",
              "                                                 random_state=1, solver='lbfgs',\n",
              "                                                 tol=0.0001, verbose=0,\n",
              "                                                 warm_start=False)),\n",
              "                             ('rf',\n",
              "                              RandomForestClassifier(bootstrap=True,\n",
              "                                                     ccp_alpha=0.0,\n",
              "                                                     class_weight=None,\n",
              "                                                     criter...\n",
              "                                                     max_samples=None,\n",
              "                                                     min_impurity_decrease=0.0,\n",
              "                                                     min_impurity_split=None,\n",
              "                                                     min_samples_leaf=1,\n",
              "                                                     min_samples_split=2,\n",
              "                                                     min_weight_fraction_leaf=0.0,\n",
              "                                                     n_estimators=100,\n",
              "                                                     n_jobs=None,\n",
              "                                                     oob_score=False,\n",
              "                                                     random_state=1, verbose=0,\n",
              "                                                     warm_start=False)),\n",
              "                             ('mnb',\n",
              "                              MultinomialNB(alpha=0.1, class_prior=None,\n",
              "                                            fit_prior=True))],\n",
              "                 flatten_transform=True, n_jobs=None, voting='hard',\n",
              "                 weights=None)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 58
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gR0c5HUhmql1",
        "colab_type": "code",
        "outputId": "4d148d6a-4fe3-403c-919d-0e568d38c2aa",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "from sklearn.metrics import  classification_report\n",
        "from sklearn.metrics import accuracy_score\n",
        "bbc_news_pred_labels = eclf1.predict(bbc_news_test_vectors)\n",
        "accuracy_score(bbc_news_test_labels, bbc_news_pred_labels) "
      ],
      "execution_count": 59,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9850299401197605"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 59
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uJnv_f1k2OJK",
        "colab_type": "text"
      },
      "source": [
        "# **STACKING CLASSIFIER**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YWkg643Dghpt",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 639
        },
        "outputId": "c3d789b9-d022-497f-cb34-5ddfc3e98645"
      },
      "source": [
        "from sklearn.ensemble import StackingClassifier\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.svm import LinearSVC\n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "clf1 = LogisticRegression(C = 10, random_state=1)\n",
        "clf2 = RandomForestClassifier(n_estimators=100,random_state=1)\n",
        "clf3 = MultinomialNB(alpha = 0.1)\n",
        "\n",
        "clf = StackingClassifier(estimators = [('lr',clf1),('rfc',clf2),('mnb',clf3)] , final_estimator = RandomForestClassifier(n_estimators=100,random_state=1) )\n",
        "clf.fit(bbc_news_train_vectors.toarray(), bbc_news_train_labels)"
      ],
      "execution_count": 67,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "StackingClassifier(cv=None,\n",
              "                   estimators=[('lr',\n",
              "                                LogisticRegression(C=10, class_weight=None,\n",
              "                                                   dual=False,\n",
              "                                                   fit_intercept=True,\n",
              "                                                   intercept_scaling=1,\n",
              "                                                   l1_ratio=None, max_iter=100,\n",
              "                                                   multi_class='auto',\n",
              "                                                   n_jobs=None, penalty='l2',\n",
              "                                                   random_state=1,\n",
              "                                                   solver='lbfgs', tol=0.0001,\n",
              "                                                   verbose=0,\n",
              "                                                   warm_start=False)),\n",
              "                               ('rfc',\n",
              "                                RandomForestClassifier(bootstrap=True,\n",
              "                                                       ccp_alpha=0.0,\n",
              "                                                       class_weight=...\n",
              "                                                          criterion='gini',\n",
              "                                                          max_depth=None,\n",
              "                                                          max_features='auto',\n",
              "                                                          max_leaf_nodes=None,\n",
              "                                                          max_samples=None,\n",
              "                                                          min_impurity_decrease=0.0,\n",
              "                                                          min_impurity_split=None,\n",
              "                                                          min_samples_leaf=1,\n",
              "                                                          min_samples_split=2,\n",
              "                                                          min_weight_fraction_leaf=0.0,\n",
              "                                                          n_estimators=100,\n",
              "                                                          n_jobs=None,\n",
              "                                                          oob_score=False,\n",
              "                                                          random_state=1,\n",
              "                                                          verbose=0,\n",
              "                                                          warm_start=False),\n",
              "                   n_jobs=None, passthrough=False, stack_method='auto',\n",
              "                   verbose=0)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 67
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FDrVRPnKLCUh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "8340ef18-807b-4457-f1cb-c166b016011c"
      },
      "source": [
        "from sklearn.metrics import  classification_report\n",
        "from sklearn.metrics import accuracy_score\n",
        "bbc_news_pred_labels = clf.predict(bbc_news_test_vectors)\n",
        "accuracy_score(bbc_news_test_labels, bbc_news_pred_labels) "
      ],
      "execution_count": 68,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9865269461077845"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 68
        }
      ]
    }
  ]
}