Savinova_HW1_news_classification.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": "3.8.2"
    },
    "colab": {
      "name": "Savinova_HW1-news_classification.ipynb",
      "provenance": []
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lgJ-KkwLBufI"
      },
      "source": [
        "В этом дз необходимо обучить модели для классификации. Cегодня попробуем отличать настоящие новости от фейковых\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RRmNaKAlBufW"
      },
      "source": [
        "# ячейка импортов\n",
        "\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "\n",
        "# импорт моделей\n",
        "from sklearn import feature_extraction, model_selection, naive_bayes, metrics, svm\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier\n",
        "\n",
        "from sklearn.metrics import classification_report # табличка с метриками\n",
        "\n",
        "from sklearn.feature_extraction.text import CountVectorizer\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "from collections import Counter\n",
        "\n",
        "import warnings\n",
        "warnings.filterwarnings(\"ignore\")\n",
        "# %matplotlib inline \n"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": "OK"
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 75
        },
        "id": "zFXKFo0xqR6C",
        "outputId": "7fe9c9c3-18f6-4ab9-f3f4-4bdc6f58c583"
      },
      "source": [
        "from google.colab import files\n",
        "uploaded = files.upload()"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-6b56fecc-f594-4872-83d5-0f633cd8f2bc\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-6b56fecc-f594-4872-83d5-0f633cd8f2bc\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving news_fake-n-true.csv.zip to news_fake-n-true.csv.zip\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "k1ZMLj-SBufa"
      },
      "source": [
        "# Exploratory Data Analysis (EDA)\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "53vORC0_Bufc",
        "outputId": "0d74470a-4d8d-4107-c95f-97bb3a0f1953"
      },
      "source": [
        "! unzip \"news_fake-n-true.csv.zip\"\n"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Archive:  news_fake-n-true.csv.zip\n",
            "replace news_fake-n-true.csv? [y]es, [n]o, [A]ll, [N]one, [r]ename: y\n",
            "  inflating: news_fake-n-true.csv    \n",
            "replace __MACOSX/._news_fake-n-true.csv? [y]es, [n]o, [A]ll, [N]one, [r]ename: y\n",
            "  inflating: __MACOSX/._news_fake-n-true.csv  \n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 175
        },
        "id": "-dGadBogBufg",
        "outputId": "2ebced02-ed2e-4bd0-9a51-3bb2dd837a47"
      },
      "source": [
        "data = pd.read_csv(\"news_fake-n-true.csv\")\n",
        "data.head(4)"
      ],
      "execution_count": 8,
      "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>title</th>\n",
              "      <th>text</th>\n",
              "      <th>subject</th>\n",
              "      <th>date</th>\n",
              "      <th>label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>ADVISORY: Story on Kushner visiting Iraq withd...</td>\n",
              "      <td>(Reuters) - The story “Trump son-in-law Jared ...</td>\n",
              "      <td>politicsNews</td>\n",
              "      <td>April 3, 2017</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>Kasich campaign denies report of Romney endors...</td>\n",
              "      <td>WASHINGTON (Reuters) - Republican presidential...</td>\n",
              "      <td>politicsNews</td>\n",
              "      <td>March 14, 2016</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>WATCH: Colbert Devises Perfect Advertisement ...</td>\n",
              "      <td>If America is going to end the day Donald Trum...</td>\n",
              "      <td>News</td>\n",
              "      <td>May 20, 2016</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>Trump says hopes for Cuban 'freedom' after Cas...</td>\n",
              "      <td>WEST PALM BEACH, Fla. (Reuters) - U.S. Preside...</td>\n",
              "      <td>politicsNews</td>\n",
              "      <td>November 26, 2016</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Unnamed: 0  ... label\n",
              "0           0  ...     1\n",
              "1           1  ...     1\n",
              "2           2  ...     0\n",
              "3           3  ...     1\n",
              "\n",
              "[4 rows x 6 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vB8rDaRrBufj"
      },
      "source": [
        "**колонка label отвечает за тип новости: настоящие имеют лейбл 1,  фейковые новости -- с лейблом 0** "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 423
        },
        "id": "MaQjeHVQBufk",
        "outputId": "d6023c54-167f-4e7d-857c-551fffed2f56"
      },
      "source": [
        "# создадим новый датафрейм из старого\n",
        "\n",
        "data = data.drop(['Unnamed: 0','title','subject','date'], axis=1)\n",
        "\n",
        "display(data)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "display_data",
          "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>(Reuters) - The story “Trump son-in-law Jared ...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>WASHINGTON (Reuters) - Republican presidential...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>If America is going to end the day Donald Trum...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>WEST PALM BEACH, Fla. (Reuters) - U.S. Preside...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Well, it seems that despite the efforts of not...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44893</th>\n",
              "      <td>Donald Trump s economic proposals would be dis...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44894</th>\n",
              "      <td>INDIANAPOLIS (Reuters) - President Donald Trum...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44895</th>\n",
              "      <td>During a private meeting of the most senior Re...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44896</th>\n",
              "      <td>AMMAN (Reuters) - Sabih al-Masri, Jordan s mos...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44897</th>\n",
              "      <td>BEIRUT (Reuters) - Lebanese President Michel A...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>44898 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                    text  label\n",
              "0      (Reuters) - The story “Trump son-in-law Jared ...      1\n",
              "1      WASHINGTON (Reuters) - Republican presidential...      1\n",
              "2      If America is going to end the day Donald Trum...      0\n",
              "3      WEST PALM BEACH, Fla. (Reuters) - U.S. Preside...      1\n",
              "4      Well, it seems that despite the efforts of not...      0\n",
              "...                                                  ...    ...\n",
              "44893  Donald Trump s economic proposals would be dis...      0\n",
              "44894  INDIANAPOLIS (Reuters) - President Donald Trum...      1\n",
              "44895  During a private meeting of the most senior Re...      0\n",
              "44896  AMMAN (Reuters) - Sabih al-Masri, Jordan s mos...      1\n",
              "44897  BEIRUT (Reuters) - Lebanese President Michel A...      1\n",
              "\n",
              "[44898 rows x 2 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IpJXbSFKBufn"
      },
      "source": [
        "распределение настоящих и фейковых новостей примерно равное"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 264
        },
        "id": "v8YLcZC8Bufq",
        "outputId": "0358efda-2785-4ded-c90e-166666535bb5"
      },
      "source": [
        "classes =pd.value_counts(data[\"label\"])\n",
        "\n",
        "classes.plot(kind = 'pie',  autopct='%0.f%%') \n",
        "plt.title(\"Classes distribution\")\n",
        "plt.ylabel('')\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KCMyYlybBufs"
      },
      "source": [
        "**давайте попробуем предсказать:** окажется ли текст настоящей новостью или фейковой\n",
        "\n",
        "- Векторизуйте тексты\n",
        "- Разделите данные на обучающую и тестовую выборку\n",
        "\n",
        "Попробуйте обучить разные модели для классификации на данных\n",
        "1. Naive Bayes\n",
        "2. SVM \n",
        "3. Decision Trees\n",
        "4. Random Forest\n",
        "5. Boosting\n",
        "\n",
        "Для каждой модели посмотрите на метрики с помощью функции [classification report](https://scikit-learn.org/stable/modules/generated/sklearn.metrics.classification_report.html).\n",
        "\n",
        "Какая/какие модель/ли справляются с задачей наиболее эффективно?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YRvuJBxbBufu",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "1aaa0a21-d597-4308-8f9b-06e6f2f5e622"
      },
      "source": [
        "# заберем все слова по условию, заджойним в строку, разобьем по пробелам, возьмем топ-20\n",
        "count1 = Counter(\" \".join(data[data['label']==1][\"text\"]).lower().split()).most_common(20) \n",
        "df1 = pd.DataFrame.from_dict(count1) # новый дф\n",
        "df1 = df1.rename(columns={0: \"words in non-fake\", 1 : \"count\"})\n",
        "\n",
        "# тот же набор операций для второго подкорпуса\n",
        "count2 = Counter(\" \".join(data[data['label']==0][\"text\"]).lower().split()).most_common(20)\n",
        "df2 = pd.DataFrame.from_dict(count2)\n",
        "df2 = df2.rename(columns={0: \"words in fake\", 1 : \"count\"})\n",
        "\n",
        "display(df1,df2)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "display_data",
          "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>words in non-fake</th>\n",
              "      <th>count</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>the</td>\n",
              "      <td>477832</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>to</td>\n",
              "      <td>244369</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>of</td>\n",
              "      <td>204654</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>a</td>\n",
              "      <td>196601</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>and</td>\n",
              "      <td>180796</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>in</td>\n",
              "      <td>179501</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>on</td>\n",
              "      <td>107739</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>that</td>\n",
              "      <td>84922</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>for</td>\n",
              "      <td>79112</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>said</td>\n",
              "      <td>72035</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>is</td>\n",
              "      <td>54706</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>with</td>\n",
              "      <td>54074</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>he</td>\n",
              "      <td>52548</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>was</td>\n",
              "      <td>47762</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>by</td>\n",
              "      <td>47484</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>as</td>\n",
              "      <td>46871</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>has</td>\n",
              "      <td>46036</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>s</td>\n",
              "      <td>44709</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>it</td>\n",
              "      <td>41937</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19</th>\n",
              "      <td>from</td>\n",
              "      <td>39189</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   words in non-fake   count\n",
              "0                the  477832\n",
              "1                 to  244369\n",
              "2                 of  204654\n",
              "3                  a  196601\n",
              "4                and  180796\n",
              "5                 in  179501\n",
              "6                 on  107739\n",
              "7               that   84922\n",
              "8                for   79112\n",
              "9               said   72035\n",
              "10                is   54706\n",
              "11              with   54074\n",
              "12                he   52548\n",
              "13               was   47762\n",
              "14                by   47484\n",
              "15                as   46871\n",
              "16               has   46036\n",
              "17                 s   44709\n",
              "18                it   41937\n",
              "19              from   39189"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "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>words in fake</th>\n",
              "      <th>count</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>the</td>\n",
              "      <td>525528</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>to</td>\n",
              "      <td>288565</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>of</td>\n",
              "      <td>235161</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>and</td>\n",
              "      <td>222329</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>a</td>\n",
              "      <td>209598</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>in</td>\n",
              "      <td>162846</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>that</td>\n",
              "      <td>144901</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>s</td>\n",
              "      <td>128331</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>is</td>\n",
              "      <td>107720</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>for</td>\n",
              "      <td>91066</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>on</td>\n",
              "      <td>79829</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>he</td>\n",
              "      <td>76896</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>it</td>\n",
              "      <td>70359</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>was</td>\n",
              "      <td>66944</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>trump</td>\n",
              "      <td>65511</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>with</td>\n",
              "      <td>62025</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>his</td>\n",
              "      <td>57885</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>as</td>\n",
              "      <td>55422</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>this</td>\n",
              "      <td>51753</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19</th>\n",
              "      <td>be</td>\n",
              "      <td>47794</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   words in fake   count\n",
              "0            the  525528\n",
              "1             to  288565\n",
              "2             of  235161\n",
              "3            and  222329\n",
              "4              a  209598\n",
              "5             in  162846\n",
              "6           that  144901\n",
              "7              s  128331\n",
              "8             is  107720\n",
              "9            for   91066\n",
              "10            on   79829\n",
              "11            he   76896\n",
              "12            it   70359\n",
              "13           was   66944\n",
              "14         trump   65511\n",
              "15          with   62025\n",
              "16           his   57885\n",
              "17            as   55422\n",
              "18          this   51753\n",
              "19            be   47794"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wObpQ7vlEd_f"
      },
      "source": [
        "f = feature_extraction.text.CountVectorizer(stop_words = 'english')"
      ],
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XxL8q9udEhSZ",
        "outputId": "3dcd7111-947c-49cf-8cbe-cc3a0895f671"
      },
      "source": [
        "data[\"text\"]"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0        (Reuters) - The story “Trump son-in-law Jared ...\n",
              "1        WASHINGTON (Reuters) - Republican presidential...\n",
              "2        If America is going to end the day Donald Trum...\n",
              "3        WEST PALM BEACH, Fla. (Reuters) - U.S. Preside...\n",
              "4        Well, it seems that despite the efforts of not...\n",
              "                               ...                        \n",
              "44893    Donald Trump s economic proposals would be dis...\n",
              "44894    INDIANAPOLIS (Reuters) - President Donald Trum...\n",
              "44895    During a private meeting of the most senior Re...\n",
              "44896    AMMAN (Reuters) - Sabih al-Masri, Jordan s mos...\n",
              "44897    BEIRUT (Reuters) - Lebanese President Michel A...\n",
              "Name: text, Length: 44898, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cw38G-vuKsCp"
      },
      "source": [
        "from sklearn.feature_extraction.text import CountVectorizer\n",
        "vectorizer = CountVectorizer()\n",
        "X = f.fit_transform(data[\"text\"])\n",
        "# print(f.get_feature_names()) # посмотрим на признаки\n",
        "\n"
      ],
      "execution_count": 52,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kzjYuv4OKzLL",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 423
        },
        "outputId": "46f79737-fb8d-4b53-cee0-5b676a4646cc"
      },
      "source": [
        "data"
      ],
      "execution_count": 53,
      "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>(Reuters) - The story “Trump son-in-law Jared ...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>WASHINGTON (Reuters) - Republican presidential...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>If America is going to end the day Donald Trum...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>WEST PALM BEACH, Fla. (Reuters) - U.S. Preside...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Well, it seems that despite the efforts of not...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44893</th>\n",
              "      <td>Donald Trump s economic proposals would be dis...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44894</th>\n",
              "      <td>INDIANAPOLIS (Reuters) - President Donald Trum...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44895</th>\n",
              "      <td>During a private meeting of the most senior Re...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44896</th>\n",
              "      <td>AMMAN (Reuters) - Sabih al-Masri, Jordan s mos...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44897</th>\n",
              "      <td>BEIRUT (Reuters) - Lebanese President Michel A...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>44898 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                    text  label\n",
              "0      (Reuters) - The story “Trump son-in-law Jared ...      1\n",
              "1      WASHINGTON (Reuters) - Republican presidential...      1\n",
              "2      If America is going to end the day Donald Trum...      0\n",
              "3      WEST PALM BEACH, Fla. (Reuters) - U.S. Preside...      1\n",
              "4      Well, it seems that despite the efforts of not...      0\n",
              "...                                                  ...    ...\n",
              "44893  Donald Trump s economic proposals would be dis...      0\n",
              "44894  INDIANAPOLIS (Reuters) - President Donald Trum...      1\n",
              "44895  During a private meeting of the most senior Re...      0\n",
              "44896  AMMAN (Reuters) - Sabih al-Masri, Jordan s mos...      1\n",
              "44897  BEIRUT (Reuters) - Lebanese President Michel A...      1\n",
              "\n",
              "[44898 rows x 2 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 53
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        },
        "id": "64znRvT5HoYE",
        "outputId": "68c265fd-4283-4139-b2f1-c86ba22fadfa"
      },
      "source": [
        "display(X)"
      ],
      "execution_count": 54,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<44898x121690 sparse matrix of type '<class 'numpy.int64'>'\n",
              "\twith 6881046 stored elements in Compressed Sparse Row format>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cVE4SpfZCIlY",
        "outputId": "67bff639-741a-4ba6-b3ef-0bd02249125c"
      },
      "source": [
        "# новые переменные\n",
        "X_train, X_test, y_train, y_test = model_selection.train_test_split(X, data['label'], \n",
        "                                                                    test_size=0.2, random_state=50)\n",
        "\n",
        "# print(X_train.shape, X_test.shape)\n",
        "sum(y_test == 1)"
      ],
      "execution_count": 55,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "4262"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 55
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        },
        "id": "FN9Q6xpNR6vO",
        "outputId": "880a2f9e-a6d2-4e46-9445-43d6e72d048d"
      },
      "source": [
        "# для проверки\n",
        "display(X_train.shape)\n",
        "# для проверки (какие уникальные значения есть)\n",
        "display(y_train.unique())"
      ],
      "execution_count": 56,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "(35918, 121690)"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "array([1, 0])"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "olH5WBEzTkNz"
      },
      "source": [
        "MultinomialNB"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wqkEBomyCd3u"
      },
      "source": [
        "list_alpha = np.arange(1/100000, 20, 0.11) \n",
        "#возьмем лист равномерно распределенных значений от 1/100000 до 20(можно выбрать другие числа,эти были выбраны экспериментально)\n",
        "\n",
        "score_train = np.zeros(len(list_alpha)) # задали пустые списки по длине изначального листа\n",
        "score_test = np.zeros(len(list_alpha))\n",
        "recall_test = np.zeros(len(list_alpha))\n",
        "precision_test= np.zeros(len(list_alpha))\n",
        "count = 0"
      ],
      "execution_count": 57,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VaWlCNKqCdrG"
      },
      "source": [
        "for alpha in list_alpha: # для каждого значения в списке альф\n",
        "    bayes = naive_bayes.MultinomialNB(alpha=alpha) # используем модель\n",
        "    bayes.fit(X_train, y_train) # тренируем модель\n",
        "    \n",
        "    score_train[count] = bayes.score(X_train, y_train) # итерируемся по индексу в листе, добавляем значения accuracy\n",
        "    \n",
        "    score_test[count]= bayes.score(X_test, y_test)\n",
        "    \n",
        "    recall_test[count] = metrics.recall_score(y_test, bayes.predict(X_test))\n",
        "    precision_test[count] = metrics.precision_score(y_test, bayes.predict(X_test))  #какие из новостей действительно фейк\n",
        "    \n",
        "    count = count + 1 # после одной итерации, делаем новый шаг"
      ],
      "execution_count": 58,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iB_uqn-jXV61",
        "outputId": "527af9be-a767-4dce-89c6-bd6fb25b2272"
      },
      "source": [
        "# насколько хорошо модель справляется с данными.\n",
        "score_train"
      ],
      "execution_count": 59,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.97853444, 0.96441895, 0.96280417, 0.96185756, 0.96102233,\n",
              "       0.96018709, 0.95957459, 0.95910129, 0.95882287, 0.95854446,\n",
              "       0.95812684, 0.95770923, 0.95745866, 0.95729161, 0.9571524 ,\n",
              "       0.95698536, 0.95684615, 0.95665126, 0.95656774, 0.95651206,\n",
              "       0.95642853, 0.95623364, 0.95598307, 0.95581603, 0.9555933 ,\n",
              "       0.95545409, 0.9552592 , 0.95512   , 0.95500863, 0.95484158,\n",
              "       0.95475806, 0.95467454, 0.9546467 , 0.95456317, 0.95436828,\n",
              "       0.9543126 , 0.95422908, 0.95414555, 0.95406203, 0.95389498,\n",
              "       0.95381146, 0.95372794, 0.95367225, 0.95367225, 0.95361657,\n",
              "       0.95364441, 0.95356089, 0.95353305, 0.95350521, 0.95347737,\n",
              "       0.953366  , 0.95333816, 0.95333816, 0.95319895, 0.95308759,\n",
              "       0.95303191, 0.95297622, 0.95292054, 0.95286486, 0.95278134,\n",
              "       0.95272565, 0.95269781, 0.95261429, 0.95258645, 0.95266997,\n",
              "       0.95266997, 0.95264213, 0.95264213, 0.95264213, 0.95258645,\n",
              "       0.95250292, 0.95250292, 0.95250292, 0.95244724, 0.95244724,\n",
              "       0.95236372, 0.95236372, 0.95233588, 0.95228019, 0.95228019,\n",
              "       0.95225235, 0.95219667, 0.95216883, 0.95214099, 0.95214099,\n",
              "       0.95208531, 0.95202962, 0.95202962, 0.95200178, 0.95200178,\n",
              "       0.95197394, 0.95180689, 0.95175121, 0.95175121, 0.95169553,\n",
              "       0.95163985, 0.95172337, 0.95163985, 0.95163985, 0.95161201,\n",
              "       0.95155632, 0.95155632, 0.95152848, 0.95152848, 0.95155632,\n",
              "       0.95158416, 0.95158416, 0.95150064, 0.95144496, 0.95144496,\n",
              "       0.95144496, 0.95141712, 0.95138928, 0.95130575, 0.95130575,\n",
              "       0.95133359, 0.95136143, 0.95133359, 0.95122223, 0.95119439,\n",
              "       0.95119439, 0.95127791, 0.95125007, 0.95122223, 0.95125007,\n",
              "       0.95116655, 0.95116655, 0.9511387 , 0.95119439, 0.9511387 ,\n",
              "       0.95111086, 0.95111086, 0.9509995 , 0.95097166, 0.95097166,\n",
              "       0.95091598, 0.95088813, 0.95083245, 0.95077677, 0.95074893,\n",
              "       0.95072109, 0.95069325, 0.95063756, 0.95063756, 0.95060972,\n",
              "       0.95060972, 0.95058188, 0.95055404, 0.9505262 , 0.9505262 ,\n",
              "       0.95049836, 0.95044267, 0.95041483, 0.95041483, 0.95035915,\n",
              "       0.95030347, 0.95024779, 0.95024779, 0.95030347, 0.95030347,\n",
              "       0.95030347, 0.95024779, 0.9501921 , 0.95016426, 0.9500529 ,\n",
              "       0.95008074, 0.95008074, 0.95008074, 0.95008074, 0.9500529 ,\n",
              "       0.94999722, 0.95010858, 0.9501921 , 0.95016426, 0.95016426,\n",
              "       0.95016426, 0.95010858, 0.95002506, 0.94999722, 0.94996937,\n",
              "       0.94996937, 0.94996937])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 59
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 381
        },
        "id": "-7DBIQ0uCdcQ",
        "outputId": "73a9fe4b-caff-44af-f4ad-788fb1216136"
      },
      "source": [
        "matrix = np.matrix(np.c_[list_alpha, score_train, score_test, recall_test, precision_test])\n",
        "models = pd.DataFrame(data = matrix, columns = \n",
        "             ['alpha', 'Train Accuracy', 'Test Accuracy', 'Test Recall', 'Test Precision'])\n",
        "\n",
        "\n",
        "display(models.head(10))\n",
        "\n",
        "display(matrix.shape)"
      ],
      "execution_count": 60,
      "outputs": [
        {
          "output_type": "display_data",
          "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>alpha</th>\n",
              "      <th>Train Accuracy</th>\n",
              "      <th>Test Accuracy</th>\n",
              "      <th>Test Recall</th>\n",
              "      <th>Test Precision</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.00001</td>\n",
              "      <td>0.978534</td>\n",
              "      <td>0.960690</td>\n",
              "      <td>0.947677</td>\n",
              "      <td>0.968817</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.11001</td>\n",
              "      <td>0.964419</td>\n",
              "      <td>0.958018</td>\n",
              "      <td>0.953778</td>\n",
              "      <td>0.957597</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.22001</td>\n",
              "      <td>0.962804</td>\n",
              "      <td>0.957906</td>\n",
              "      <td>0.954012</td>\n",
              "      <td>0.957156</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.33001</td>\n",
              "      <td>0.961858</td>\n",
              "      <td>0.957461</td>\n",
              "      <td>0.953778</td>\n",
              "      <td>0.956471</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.44001</td>\n",
              "      <td>0.961022</td>\n",
              "      <td>0.956793</td>\n",
              "      <td>0.953778</td>\n",
              "      <td>0.955122</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>0.55001</td>\n",
              "      <td>0.960187</td>\n",
              "      <td>0.956347</td>\n",
              "      <td>0.953543</td>\n",
              "      <td>0.954439</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>0.66001</td>\n",
              "      <td>0.959575</td>\n",
              "      <td>0.955679</td>\n",
              "      <td>0.953074</td>\n",
              "      <td>0.953521</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>0.77001</td>\n",
              "      <td>0.959101</td>\n",
              "      <td>0.955345</td>\n",
              "      <td>0.953074</td>\n",
              "      <td>0.952850</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>0.88001</td>\n",
              "      <td>0.958823</td>\n",
              "      <td>0.955011</td>\n",
              "      <td>0.953308</td>\n",
              "      <td>0.951968</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>0.99001</td>\n",
              "      <td>0.958544</td>\n",
              "      <td>0.954120</td>\n",
              "      <td>0.952604</td>\n",
              "      <td>0.950820</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     alpha  Train Accuracy  Test Accuracy  Test Recall  Test Precision\n",
              "0  0.00001        0.978534       0.960690     0.947677        0.968817\n",
              "1  0.11001        0.964419       0.958018     0.953778        0.957597\n",
              "2  0.22001        0.962804       0.957906     0.954012        0.957156\n",
              "3  0.33001        0.961858       0.957461     0.953778        0.956471\n",
              "4  0.44001        0.961022       0.956793     0.953778        0.955122\n",
              "5  0.55001        0.960187       0.956347     0.953543        0.954439\n",
              "6  0.66001        0.959575       0.955679     0.953074        0.953521\n",
              "7  0.77001        0.959101       0.955345     0.953074        0.952850\n",
              "8  0.88001        0.958823       0.955011     0.953308        0.951968\n",
              "9  0.99001        0.958544       0.954120     0.952604        0.950820"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "(182, 5)"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 129
        },
        "id": "TI0C-D1mC8MQ",
        "outputId": "2eb2bbb3-e111-4233-8481-763196b4a815"
      },
      "source": [
        "#модель с наивысшей test_precision\n",
        "best_index = models['Test Precision'].idxmax()\n",
        "best_index\n",
        "\n",
        "# display(best_index)\n",
        "\n",
        "display(models.iloc[best_index])"
      ],
      "execution_count": 61,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "alpha             0.000010\n",
              "Train Accuracy    0.978534\n",
              "Test Accuracy     0.960690\n",
              "Test Recall       0.947677\n",
              "Test Precision    0.968817\n",
              "Name: 0, dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 423
        },
        "id": "Rg3g2b_7DBVW",
        "outputId": "154ce819-2766-4acc-aaba-c4d523595291"
      },
      "source": [
        "models[models['Test Recall']>0.9].sort_values(by=\"Test Precision\", ascending = False)\n",
        "# проверить модели с высокими знаечниями recall и precision."
      ],
      "execution_count": 62,
      "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>alpha</th>\n",
              "      <th>Train Accuracy</th>\n",
              "      <th>Test Accuracy</th>\n",
              "      <th>Test Recall</th>\n",
              "      <th>Test Precision</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.00001</td>\n",
              "      <td>0.978534</td>\n",
              "      <td>0.960690</td>\n",
              "      <td>0.947677</td>\n",
              "      <td>0.968817</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.11001</td>\n",
              "      <td>0.964419</td>\n",
              "      <td>0.958018</td>\n",
              "      <td>0.953778</td>\n",
              "      <td>0.957597</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.22001</td>\n",
              "      <td>0.962804</td>\n",
              "      <td>0.957906</td>\n",
              "      <td>0.954012</td>\n",
              "      <td>0.957156</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.33001</td>\n",
              "      <td>0.961858</td>\n",
              "      <td>0.957461</td>\n",
              "      <td>0.953778</td>\n",
              "      <td>0.956471</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.44001</td>\n",
              "      <td>0.961022</td>\n",
              "      <td>0.956793</td>\n",
              "      <td>0.953778</td>\n",
              "      <td>0.955122</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>59</th>\n",
              "      <td>6.49001</td>\n",
              "      <td>0.952781</td>\n",
              "      <td>0.949666</td>\n",
              "      <td>0.949085</td>\n",
              "      <td>0.945093</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>63</th>\n",
              "      <td>6.93001</td>\n",
              "      <td>0.952586</td>\n",
              "      <td>0.949555</td>\n",
              "      <td>0.948850</td>\n",
              "      <td>0.945081</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>62</th>\n",
              "      <td>6.82001</td>\n",
              "      <td>0.952614</td>\n",
              "      <td>0.949555</td>\n",
              "      <td>0.948850</td>\n",
              "      <td>0.945081</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>45</th>\n",
              "      <td>4.95001</td>\n",
              "      <td>0.953644</td>\n",
              "      <td>0.950111</td>\n",
              "      <td>0.950258</td>\n",
              "      <td>0.944937</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44</th>\n",
              "      <td>4.84001</td>\n",
              "      <td>0.953617</td>\n",
              "      <td>0.950111</td>\n",
              "      <td>0.950258</td>\n",
              "      <td>0.944937</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>182 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      alpha  Train Accuracy  Test Accuracy  Test Recall  Test Precision\n",
              "0   0.00001        0.978534       0.960690     0.947677        0.968817\n",
              "1   0.11001        0.964419       0.958018     0.953778        0.957597\n",
              "2   0.22001        0.962804       0.957906     0.954012        0.957156\n",
              "3   0.33001        0.961858       0.957461     0.953778        0.956471\n",
              "4   0.44001        0.961022       0.956793     0.953778        0.955122\n",
              "..      ...             ...            ...          ...             ...\n",
              "59  6.49001        0.952781       0.949666     0.949085        0.945093\n",
              "63  6.93001        0.952586       0.949555     0.948850        0.945081\n",
              "62  6.82001        0.952614       0.949555     0.948850        0.945081\n",
              "45  4.95001        0.953644       0.950111     0.950258        0.944937\n",
              "44  4.84001        0.953617       0.950111     0.950258        0.944937\n",
              "\n",
              "[182 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 62
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RhwvpXMhDKqa",
        "outputId": "26dad632-5e14-47d3-d3cc-8d6fe2024d70"
      },
      "source": [
        "best_index = models[models['Test Precision']>0.95]['Test Accuracy'].idxmax()\n",
        "\n",
        "models.iloc[best_index]"
      ],
      "execution_count": 64,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "alpha             0.000010\n",
              "Train Accuracy    0.978534\n",
              "Test Accuracy     0.960690\n",
              "Test Recall       0.947677\n",
              "Test Precision    0.968817\n",
              "Name: 0, dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 64
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 423
        },
        "id": "DPX48wiXDeFy",
        "outputId": "769d0551-05f5-4d64-bba7-886090d420cd"
      },
      "source": [
        "bayes = naive_bayes.MultinomialNB(alpha= 0.000010) # используем лучшую модель\n",
        "bayes.fit(X_train, y_train) # тренируем алгоритм на данных\n",
        "\n",
        "bayes.predict(X_test) # метод для предсказания\n",
        "\n",
        "# сделаем датафрейм с колонками для настоящего класса и для предсказанного\n",
        "\n",
        "new_df = pd.DataFrame(data = y_test)\n",
        "new_df[\"predicted\"]=bayes.predict(X_test) # добавляем новую колонку с предсказаниями\n",
        "\n",
        "new_df"
      ],
      "execution_count": 65,
      "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>label</th>\n",
              "      <th>predicted</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>18322</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6805</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11776</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>43603</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1166</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3814</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>27520</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11071</th>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2479</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10428</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>8980 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "       label  predicted\n",
              "18322      0          0\n",
              "6805       0          0\n",
              "11776      0          0\n",
              "43603      1          1\n",
              "1166       1          1\n",
              "...      ...        ...\n",
              "3814       0          0\n",
              "27520      1          1\n",
              "11071      0          1\n",
              "2479       0          0\n",
              "10428      1          0\n",
              "\n",
              "[8980 rows x 2 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 65
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aMGWlaFmeCDe",
        "outputId": "146a572c-925a-408e-ea0a-d07617b2662e"
      },
      "source": [
        "# classification_report\n",
        "report = classification_report(y_test, bayes.predict(X_test), target_names=['Non-fake', 'Fake'])\n",
        "print(report)"
      ],
      "execution_count": 66,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "    Non-fake       0.95      0.97      0.96      4718\n",
            "        Fake       0.97      0.95      0.96      4262\n",
            "\n",
            "    accuracy                           0.96      8980\n",
            "   macro avg       0.96      0.96      0.96      8980\n",
            "weighted avg       0.96      0.96      0.96      8980\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6czcVy5XcxGg"
      },
      "source": [
        "SVM"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C-6tFTb4dB_E"
      },
      "source": [
        "list_C = np.arange(500, 2000, 100) #100000\n",
        "\n",
        "score_train = np.zeros(len(list_C))\n",
        "score_test = np.zeros(len(list_C))\n",
        "recall_test = np.zeros(len(list_C))\n",
        "precision_test= np.zeros(len(list_C))\n",
        "count = 0\n",
        "for C in list_C:\n",
        "    svc = svm.SVC(C=C)\n",
        "    svc.fit(X_train, y_train)\n",
        "    score_train[count] = svc.score(X_train, y_train)\n",
        "    score_test[count]= svc.score(X_test, y_test)\n",
        "    recall_test[count] = metrics.recall_score(y_test, svc.predict(X_test))\n",
        "    precision_test[count] = metrics.precision_score(y_test, svc.predict(X_test))\n",
        "    count = count + 1"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8bmK15yCikWX"
      },
      "source": [
        "#10 Первых моделей и их метрики\n",
        "\n",
        "matrix = np.matrix(np.c_[list_C, score_train, score_test, recall_test, precision_test])\n",
        "models = pd.DataFrame(data = matrix, columns = \n",
        "             ['C', 'Train Accuracy', 'Test Accuracy', 'Test Recall', 'Test Precision'])\n",
        "models.head(n=10)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "X5zQnvw7ip6n"
      },
      "source": [
        "#выберем модель с лучшим test precision\n",
        "best_index = models['Test Precision'].idxmax()\n",
        "\n",
        "display(models.iloc[best_index, :])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YQRPGd3Tiu2O"
      },
      "source": [
        "# посмотрим на модели с почти 100% Precision\n",
        "\n",
        "models[models['Test Precision']>0.95].head(5)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oPS-Dx7miwuR"
      },
      "source": [
        "best_index = models[models['Test Precision']>0.99]['Test Accuracy'].idxmax()\n",
        "svc = svm.SVC(C=list_C[best_index])\n",
        "svc.fit(X_train, y_train)\n",
        "models.iloc[best_index, :]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JVyBEdIMiymJ"
      },
      "source": [
        "m_confusion_test = metrics.confusion_matrix(y_test, svc.predict(X_test))\n",
        "pd.DataFrame(data = m_confusion_test, columns = ['Predicted 0', 'Predicted 1'],\n",
        "            index = ['Actual 0', 'Actual 1'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SOIgg1gRNb2b"
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "# from sklearn.datasets import load_wine\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "import seaborn as sns"
      ],
      "execution_count": 67,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 423
        },
        "id": "S5boBRj_R0wn",
        "outputId": "c774403c-1a98-4a76-f259-302e39e9080b"
      },
      "source": [
        "display(data)"
      ],
      "execution_count": 68,
      "outputs": [
        {
          "output_type": "display_data",
          "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>(Reuters) - The story “Trump son-in-law Jared ...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>WASHINGTON (Reuters) - Republican presidential...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>If America is going to end the day Donald Trum...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>WEST PALM BEACH, Fla. (Reuters) - U.S. Preside...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Well, it seems that despite the efforts of not...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44893</th>\n",
              "      <td>Donald Trump s economic proposals would be dis...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44894</th>\n",
              "      <td>INDIANAPOLIS (Reuters) - President Donald Trum...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44895</th>\n",
              "      <td>During a private meeting of the most senior Re...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44896</th>\n",
              "      <td>AMMAN (Reuters) - Sabih al-Masri, Jordan s mos...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44897</th>\n",
              "      <td>BEIRUT (Reuters) - Lebanese President Michel A...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>44898 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                    text  label\n",
              "0      (Reuters) - The story “Trump son-in-law Jared ...      1\n",
              "1      WASHINGTON (Reuters) - Republican presidential...      1\n",
              "2      If America is going to end the day Donald Trum...      0\n",
              "3      WEST PALM BEACH, Fla. (Reuters) - U.S. Preside...      1\n",
              "4      Well, it seems that despite the efforts of not...      0\n",
              "...                                                  ...    ...\n",
              "44893  Donald Trump s economic proposals would be dis...      0\n",
              "44894  INDIANAPOLIS (Reuters) - President Donald Trum...      1\n",
              "44895  During a private meeting of the most senior Re...      0\n",
              "44896  AMMAN (Reuters) - Sabih al-Masri, Jordan s mos...      1\n",
              "44897  BEIRUT (Reuters) - Lebanese President Michel A...      1\n",
              "\n",
              "[44898 rows x 2 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nuNP-48aSEo7"
      },
      "source": [
        "# векторизация\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "vectorizer = TfidfVectorizer(stop_words=\"english\")\n",
        "\n",
        "X = vectorizer.fit_transform(data['text'])"
      ],
      "execution_count": 69,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AkMQ2V7BTWP_"
      },
      "source": [
        "# разделение на обучающую и тестовую выборку\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split( X, data[\"label\"], test_size=0.3, random_state=25)"
      ],
      "execution_count": 70,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2or-fTeGNZn3"
      },
      "source": [
        "Decision Tree"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S_sLpuxmiajg"
      },
      "source": [
        "# сначала обучим одно решающее дерево\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.metrics import classification_report"
      ],
      "execution_count": 71,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AbGVj6_tikAL"
      },
      "source": [
        "tree_clf = DecisionTreeClassifier()"
      ],
      "execution_count": 72,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MpdV8_MQiopp",
        "outputId": "e5a13e5a-9b09-4240-a190-82ad9f1bbfd7"
      },
      "source": [
        "tree_clf.fit(X_train, y_train)"
      ],
      "execution_count": 73,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "DecisionTreeClassifier(ccp_alpha=0.0, class_weight=None, criterion='gini',\n",
              "                       max_depth=None, max_features=None, max_leaf_nodes=None,\n",
              "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                       min_samples_leaf=1, min_samples_split=2,\n",
              "                       min_weight_fraction_leaf=0.0, presort='deprecated',\n",
              "                       random_state=None, splitter='best')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 73
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JAGiUP3OitZP"
      },
      "source": [
        "tree_y_pred = tree_clf.predict(X_test)"
      ],
      "execution_count": 74,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hIpKiJgeiuHi",
        "outputId": "223526aa-9961-4937-b29f-2bfe3ed650af"
      },
      "source": [
        "tree_y_pred"
      ],
      "execution_count": 75,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 0, 1, ..., 0, 0, 0])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 75
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Y6XrWNJMiwcI",
        "outputId": "61256857-8ef3-4719-df2e-0b0b1f5cfc9a"
      },
      "source": [
        "print(classification_report(y_test, tree_y_pred))\n"
      ],
      "execution_count": 76,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      1.00      1.00      7094\n",
            "           1       1.00      0.99      0.99      6376\n",
            "\n",
            "    accuracy                           1.00     13470\n",
            "   macro avg       1.00      1.00      1.00     13470\n",
            "weighted avg       1.00      1.00      1.00     13470\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CpUrpexeknn8"
      },
      "source": [
        "Random forest\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Tpz8cPzAkr-d"
      },
      "source": [
        "from sklearn.ensemble import RandomForestClassifier \n",
        "#forest = RandomForestClassifier(n_estimators=500, verbose=1)\n",
        "# forest = RandomForestClassifier(n_estimators=3, max_depth=5, verbose=5) #так было в лекции\n",
        "forest = RandomForestClassifier(n_estimators=200, max_depth=50, verbose=5) #так было в лекции"
      ],
      "execution_count": 77,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h73Pg6DGkzzg"
      },
      "source": [
        "используемые параметры модели: \n",
        "- n_estimators : число деревьев в ансамбле (лесе)\n",
        "- max_depth : глубина дерева \n",
        "- verbose : пояснения о процессе обучения \n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tj3y3czSlQB2",
        "outputId": "6d399efd-4822-4176-dca8-bca0787c6742"
      },
      "source": [
        "#обучим\n",
        "forest.fit(X_train, y_train)"
      ],
      "execution_count": 78,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "building tree 1 of 200\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
            "[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.5s remaining:    0.0s\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "building tree 2 of 200\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    1.0s remaining:    0.0s\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "building tree 3 of 200\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    1.5s remaining:    0.0s\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "building tree 4 of 200\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "[Parallel(n_jobs=1)]: Done   4 out of   4 | elapsed:    2.0s remaining:    0.0s\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "building tree 5 of 200\n",
            "building tree 6 of 200\n",
            "building tree 7 of 200\n",
            "building tree 8 of 200\n",
            "building tree 9 of 200\n",
            "building tree 10 of 200\n",
            "building tree 11 of 200\n",
            "building tree 12 of 200\n",
            "building tree 13 of 200\n",
            "building tree 14 of 200\n",
            "building tree 15 of 200\n",
            "building tree 16 of 200\n",
            "building tree 17 of 200\n",
            "building tree 18 of 200\n",
            "building tree 19 of 200\n",
            "building tree 20 of 200\n",
            "building tree 21 of 200\n",
            "building tree 22 of 200\n",
            "building tree 23 of 200\n",
            "building tree 24 of 200\n",
            "building tree 25 of 200\n",
            "building tree 26 of 200\n",
            "building tree 27 of 200\n",
            "building tree 28 of 200\n",
            "building tree 29 of 200\n",
            "building tree 30 of 200\n",
            "building tree 31 of 200\n",
            "building tree 32 of 200\n",
            "building tree 33 of 200\n",
            "building tree 34 of 200\n",
            "building tree 35 of 200\n",
            "building tree 36 of 200\n",
            "building tree 37 of 200\n",
            "building tree 38 of 200\n",
            "building tree 39 of 200\n",
            "building tree 40 of 200\n",
            "building tree 41 of 200\n",
            "building tree 42 of 200\n",
            "building tree 43 of 200\n",
            "building tree 44 of 200\n",
            "building tree 45 of 200\n",
            "building tree 46 of 200\n",
            "building tree 47 of 200\n",
            "building tree 48 of 200\n",
            "building tree 49 of 200\n",
            "building tree 50 of 200\n",
            "building tree 51 of 200\n",
            "building tree 52 of 200\n",
            "building tree 53 of 200\n",
            "building tree 54 of 200\n",
            "building tree 55 of 200\n",
            "building tree 56 of 200\n",
            "building tree 57 of 200\n",
            "building tree 58 of 200\n",
            "building tree 59 of 200\n",
            "building tree 60 of 200\n",
            "building tree 61 of 200\n",
            "building tree 62 of 200\n",
            "building tree 63 of 200\n",
            "building tree 64 of 200\n",
            "building tree 65 of 200\n",
            "building tree 66 of 200\n",
            "building tree 67 of 200\n",
            "building tree 68 of 200\n",
            "building tree 69 of 200\n",
            "building tree 70 of 200\n",
            "building tree 71 of 200\n",
            "building tree 72 of 200\n",
            "building tree 73 of 200\n",
            "building tree 74 of 200\n",
            "building tree 75 of 200\n",
            "building tree 76 of 200\n",
            "building tree 77 of 200\n",
            "building tree 78 of 200\n",
            "building tree 79 of 200\n",
            "building tree 80 of 200\n",
            "building tree 81 of 200\n",
            "building tree 82 of 200\n",
            "building tree 83 of 200\n",
            "building tree 84 of 200\n",
            "building tree 85 of 200\n",
            "building tree 86 of 200\n",
            "building tree 87 of 200\n",
            "building tree 88 of 200\n",
            "building tree 89 of 200\n",
            "building tree 90 of 200\n",
            "building tree 91 of 200\n",
            "building tree 92 of 200\n",
            "building tree 93 of 200\n",
            "building tree 94 of 200\n",
            "building tree 95 of 200\n",
            "building tree 96 of 200\n",
            "building tree 97 of 200\n",
            "building tree 98 of 200\n",
            "building tree 99 of 200\n",
            "building tree 100 of 200\n",
            "building tree 101 of 200\n",
            "building tree 102 of 200\n",
            "building tree 103 of 200\n",
            "building tree 104 of 200\n",
            "building tree 105 of 200\n",
            "building tree 106 of 200\n",
            "building tree 107 of 200\n",
            "building tree 108 of 200\n",
            "building tree 109 of 200\n",
            "building tree 110 of 200\n",
            "building tree 111 of 200\n",
            "building tree 112 of 200\n",
            "building tree 113 of 200\n",
            "building tree 114 of 200\n",
            "building tree 115 of 200\n",
            "building tree 116 of 200\n",
            "building tree 117 of 200\n",
            "building tree 118 of 200\n",
            "building tree 119 of 200\n",
            "building tree 120 of 200\n",
            "building tree 121 of 200\n",
            "building tree 122 of 200\n",
            "building tree 123 of 200\n",
            "building tree 124 of 200\n",
            "building tree 125 of 200\n",
            "building tree 126 of 200\n",
            "building tree 127 of 200\n",
            "building tree 128 of 200\n",
            "building tree 129 of 200\n",
            "building tree 130 of 200\n",
            "building tree 131 of 200\n",
            "building tree 132 of 200\n",
            "building tree 133 of 200\n",
            "building tree 134 of 200\n",
            "building tree 135 of 200\n",
            "building tree 136 of 200\n",
            "building tree 137 of 200\n",
            "building tree 138 of 200\n",
            "building tree 139 of 200\n",
            "building tree 140 of 200\n",
            "building tree 141 of 200\n",
            "building tree 142 of 200\n",
            "building tree 143 of 200\n",
            "building tree 144 of 200\n",
            "building tree 145 of 200\n",
            "building tree 146 of 200\n",
            "building tree 147 of 200\n",
            "building tree 148 of 200\n",
            "building tree 149 of 200\n",
            "building tree 150 of 200\n",
            "building tree 151 of 200\n",
            "building tree 152 of 200\n",
            "building tree 153 of 200\n",
            "building tree 154 of 200\n",
            "building tree 155 of 200\n",
            "building tree 156 of 200\n",
            "building tree 157 of 200\n",
            "building tree 158 of 200\n",
            "building tree 159 of 200\n",
            "building tree 160 of 200\n",
            "building tree 161 of 200\n",
            "building tree 162 of 200\n",
            "building tree 163 of 200\n",
            "building tree 164 of 200\n",
            "building tree 165 of 200\n",
            "building tree 166 of 200\n",
            "building tree 167 of 200\n",
            "building tree 168 of 200\n",
            "building tree 169 of 200\n",
            "building tree 170 of 200\n",
            "building tree 171 of 200\n",
            "building tree 172 of 200\n",
            "building tree 173 of 200\n",
            "building tree 174 of 200\n",
            "building tree 175 of 200\n",
            "building tree 176 of 200\n",
            "building tree 177 of 200\n",
            "building tree 178 of 200\n",
            "building tree 179 of 200\n",
            "building tree 180 of 200\n",
            "building tree 181 of 200\n",
            "building tree 182 of 200\n",
            "building tree 183 of 200\n",
            "building tree 184 of 200\n",
            "building tree 185 of 200\n",
            "building tree 186 of 200\n",
            "building tree 187 of 200\n",
            "building tree 188 of 200\n",
            "building tree 189 of 200\n",
            "building tree 190 of 200\n",
            "building tree 191 of 200\n",
            "building tree 192 of 200\n",
            "building tree 193 of 200\n",
            "building tree 194 of 200\n",
            "building tree 195 of 200\n",
            "building tree 196 of 200\n",
            "building tree 197 of 200\n",
            "building tree 198 of 200\n",
            "building tree 199 of 200\n",
            "building tree 200 of 200\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:  1.7min finished\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "RandomForestClassifier(bootstrap=True, ccp_alpha=0.0, class_weight=None,\n",
              "                       criterion='gini', max_depth=50, max_features='auto',\n",
              "                       max_leaf_nodes=None, max_samples=None,\n",
              "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                       min_samples_leaf=1, min_samples_split=2,\n",
              "                       min_weight_fraction_leaf=0.0, n_estimators=200,\n",
              "                       n_jobs=None, oob_score=False, random_state=None,\n",
              "                       verbose=5, warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 78
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BcI764R9lTYn",
        "outputId": "4f64067e-0971-4abf-d274-a5423be259cf"
      },
      "source": [
        "forest_y_pred = forest.predict(X_test)"
      ],
      "execution_count": 79,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
            "[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s remaining:    0.0s\n",
            "[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.0s remaining:    0.0s\n",
            "[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.0s remaining:    0.0s\n",
            "[Parallel(n_jobs=1)]: Done   4 out of   4 | elapsed:    0.0s remaining:    0.0s\n",
            "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    1.7s finished\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YDOdR3O4lbPv",
        "outputId": "7d5a0bc7-bbf6-499c-98b0-c9002149479e"
      },
      "source": [
        "print(classification_report(y_test, forest_y_pred))\n"
      ],
      "execution_count": 80,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      0.99      0.99      7094\n",
            "           1       0.99      0.99      0.99      6376\n",
            "\n",
            "    accuracy                           0.99     13470\n",
            "   macro avg       0.99      0.99      0.99     13470\n",
            "weighted avg       0.99      0.99      0.99     13470\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VXHy33MdpcJH"
      },
      "source": [
        "Boosting"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vPlF0OJNpeOZ"
      },
      "source": [
        "from sklearn.ensemble import GradientBoostingClassifier"
      ],
      "execution_count": 81,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-m0a9EIKqQs_"
      },
      "source": [
        "gb_clf = GradientBoostingClassifier(n_estimators=200, random_state=10, verbose=1) #на 100 итераций уже хорошо работает\n",
        "# loss - количество ошибок (должно уменьшаться)"
      ],
      "execution_count": 82,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HAxMqDBZqURG"
      },
      "source": [
        "используемые параметры [модели](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingClassifier.html#sklearn.ensemble.GradientBoostingClassifier):\n",
        "- loss : The loss function to be optimized. ‘deviance’ refers to deviance (= logistic regression) for classification with probabilistic outputs. For loss ‘exponential’ gradient boosting recovers the AdaBoost algorithm.\n",
        "- n_estimators : The number of boosting stages to perform. Gradient boosting is fairly robust to over-fitting so a large number usually results in better performance.\n",
        "- random_state : Controls the random seed given to each Tree estimator at each boosting iteration. In addition, it controls the random permutation of the features at each split (see Notes for more details). It also controls the random spliting of the training data to obtain a validation set if n_iter_no_change is not None. Pass an int for reproducible output across multiple function calls. See Glossary.\n",
        "- verbose : Enable verbose output. If 1 then it prints progress and performance once in a while (the more trees the lower the frequency). If greater than 1 then it prints progress and performance for every tree.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xWYwex_0qYc-",
        "outputId": "e5fbe4bb-5c37-4544-8acb-d6c0393faca7"
      },
      "source": [
        "gb_clf.fit(X_train, y_train)"
      ],
      "execution_count": 83,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "      Iter       Train Loss   Remaining Time \n",
            "         1           1.1983           10.25m\n",
            "         2           1.0459           10.16m\n",
            "         3           0.9190           10.07m\n",
            "         4           0.8118            9.96m\n",
            "         5           0.7203            9.90m\n",
            "         6           0.6414            9.84m\n",
            "         7           0.5731            9.78m\n",
            "         8           0.5137            9.73m\n",
            "         9           0.4616            9.69m\n",
            "        10           0.4159            9.63m\n",
            "        20           0.1658            9.10m\n",
            "        30           0.0833            8.60m\n",
            "        40           0.0541            8.08m\n",
            "        50           0.0399            7.55m\n",
            "        60           0.0355            7.03m\n",
            "        70           0.0274            6.51m\n",
            "        80           0.0247            6.00m\n",
            "        90           0.0234            5.50m\n",
            "       100           0.0220            4.99m\n",
            "       200           0.0157            0.00s\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "GradientBoostingClassifier(ccp_alpha=0.0, criterion='friedman_mse', init=None,\n",
              "                           learning_rate=0.1, loss='deviance', max_depth=3,\n",
              "                           max_features=None, max_leaf_nodes=None,\n",
              "                           min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                           min_samples_leaf=1, min_samples_split=2,\n",
              "                           min_weight_fraction_leaf=0.0, n_estimators=200,\n",
              "                           n_iter_no_change=None, presort='deprecated',\n",
              "                           random_state=10, subsample=1.0, tol=0.0001,\n",
              "                           validation_fraction=0.1, verbose=1,\n",
              "                           warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 83
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EV1plzk5qpK1"
      },
      "source": [
        "gb_y_pred = gb_clf.predict(X_test)"
      ],
      "execution_count": 84,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "w5BPal6qqrHB",
        "outputId": "51115ee7-417f-482a-db1f-4ce941988b41"
      },
      "source": [
        "print(classification_report(y_test, gb_y_pred))"
      ],
      "execution_count": 85,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       1.00      0.99      1.00      7094\n",
            "           1       0.99      1.00      0.99      6376\n",
            "\n",
            "    accuracy                           1.00     13470\n",
            "   macro avg       1.00      1.00      1.00     13470\n",
            "weighted avg       1.00      1.00      1.00     13470\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZBHNpJl5uRSg"
      },
      "source": [
        "Итоги: из 4-х моделей наихудшиие результаты показал Naive Bayes, однако сами результаты довольно неплохие: самое низкое значение - precision и recall у нефейковых новостей (0,95), при это f1 и accuracy - 0,96.\n",
        "Наилучшие результаты показал Boosting. Почти все значения равны единице, включая accuracy и f1. Значит модель правильно находит то, что нужно и не находит ненужное."
      ]
    }
  ]
}