Model Training Attempt 1 & 2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "51074891",
   "metadata": {},
   "source": [
    "# Import required libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "293ca3c0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd# Import cleaned dataset\n",
    "\n",
    "import seaborn as sb\n",
    "sb.set(font_scale=2) # set the default Seaborn style for graphics\n",
    "import matplotlib.pyplot as plt # we only need pyplot\n",
    "\n",
    "# For model training\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from sklearn.tree import DecisionTreeClassifier, plot_tree\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import GridSearchCV, train_test_split, StratifiedKFold, cross_val_score\n",
    "\n",
    "# For model evaluation\n",
    "from sklearn.metrics import accuracy_score, confusion_matrix, classification_report"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "afc5aa1d",
   "metadata": {},
   "source": [
    "# Import cleaned dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "7624e176",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load data back for EDA\n",
    "news_data = pd.read_csv(\"cleaned_news_data.csv\")\n",
    "news_data.set_index(\"id\", inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "0af9504d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(20584, 29)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "1fa1a027",
   "metadata": {},
   "outputs": [
    {
     "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>title</th>\n",
       "      <th>author</th>\n",
       "      <th>text</th>\n",
       "      <th>label</th>\n",
       "      <th>label_translated</th>\n",
       "      <th>title_charcount</th>\n",
       "      <th>text_charcount</th>\n",
       "      <th>title_wordcount</th>\n",
       "      <th>text_wordcount</th>\n",
       "      <th>stopwords_count_text</th>\n",
       "      <th>...</th>\n",
       "      <th>title_pos_noun</th>\n",
       "      <th>title_pos_num</th>\n",
       "      <th>title_pos_pron</th>\n",
       "      <th>title_pos_verb</th>\n",
       "      <th>text_pos_adj</th>\n",
       "      <th>text_pos_adv</th>\n",
       "      <th>text_pos_noun</th>\n",
       "      <th>text_pos_num</th>\n",
       "      <th>text_pos_pron</th>\n",
       "      <th>text_pos_verb</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>hous dem aid even see comey letter jason chaff...</td>\n",
       "      <td>Darrell Lucus</td>\n",
       "      <td>hous dem aid even see comey letter jason chaff...</td>\n",
       "      <td>1</td>\n",
       "      <td>fake</td>\n",
       "      <td>81</td>\n",
       "      <td>4930</td>\n",
       "      <td>14</td>\n",
       "      <td>820</td>\n",
       "      <td>406</td>\n",
       "      <td>...</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>102</td>\n",
       "      <td>16</td>\n",
       "      <td>234</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>flynn hillari clinton big woman campu breitbart</td>\n",
       "      <td>Daniel J. Flynn</td>\n",
       "      <td>ever get feel life circl roundabout rather hea...</td>\n",
       "      <td>0</td>\n",
       "      <td>not fake</td>\n",
       "      <td>55</td>\n",
       "      <td>4160</td>\n",
       "      <td>9</td>\n",
       "      <td>710</td>\n",
       "      <td>330</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>81</td>\n",
       "      <td>13</td>\n",
       "      <td>208</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>truth might get fire</td>\n",
       "      <td>Consortiumnews.com</td>\n",
       "      <td>truth might get fire octob tension intellig an...</td>\n",
       "      <td>1</td>\n",
       "      <td>fake</td>\n",
       "      <td>33</td>\n",
       "      <td>7692</td>\n",
       "      <td>7</td>\n",
       "      <td>1266</td>\n",
       "      <td>575</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>155</td>\n",
       "      <td>33</td>\n",
       "      <td>385</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>92</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>civilian kill singl us airstrik identifi</td>\n",
       "      <td>Jessica Purkiss</td>\n",
       "      <td>video civilian kill singl us airstrik identifi...</td>\n",
       "      <td>1</td>\n",
       "      <td>fake</td>\n",
       "      <td>63</td>\n",
       "      <td>3237</td>\n",
       "      <td>10</td>\n",
       "      <td>557</td>\n",
       "      <td>244</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>74</td>\n",
       "      <td>5</td>\n",
       "      <td>158</td>\n",
       "      <td>6</td>\n",
       "      <td>12</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>iranian woman jail fiction unpublish stori wom...</td>\n",
       "      <td>Howard Portnoy</td>\n",
       "      <td>print iranian woman sentenc six year prison ir...</td>\n",
       "      <td>1</td>\n",
       "      <td>fake</td>\n",
       "      <td>93</td>\n",
       "      <td>938</td>\n",
       "      <td>14</td>\n",
       "      <td>154</td>\n",
       "      <td>64</td>\n",
       "      <td>...</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>13</td>\n",
       "      <td>5</td>\n",
       "      <td>54</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 29 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                title              author  \\\n",
       "id                                                                          \n",
       "0   hous dem aid even see comey letter jason chaff...       Darrell Lucus   \n",
       "1     flynn hillari clinton big woman campu breitbart     Daniel J. Flynn   \n",
       "2                                truth might get fire  Consortiumnews.com   \n",
       "3            civilian kill singl us airstrik identifi     Jessica Purkiss   \n",
       "4   iranian woman jail fiction unpublish stori wom...      Howard Portnoy   \n",
       "\n",
       "                                                 text  label label_translated  \\\n",
       "id                                                                              \n",
       "0   hous dem aid even see comey letter jason chaff...      1             fake   \n",
       "1   ever get feel life circl roundabout rather hea...      0         not fake   \n",
       "2   truth might get fire octob tension intellig an...      1             fake   \n",
       "3   video civilian kill singl us airstrik identifi...      1             fake   \n",
       "4   print iranian woman sentenc six year prison ir...      1             fake   \n",
       "\n",
       "    title_charcount  text_charcount  title_wordcount  text_wordcount  \\\n",
       "id                                                                     \n",
       "0                81            4930               14             820   \n",
       "1                55            4160                9             710   \n",
       "2                33            7692                7            1266   \n",
       "3                63            3237               10             557   \n",
       "4                93             938               14             154   \n",
       "\n",
       "    stopwords_count_text  ...  title_pos_noun title_pos_num  title_pos_pron  \\\n",
       "id                        ...                                                 \n",
       "0                    406  ...               6             0               0   \n",
       "1                    330  ...               5             0               0   \n",
       "2                    575  ...               2             0               0   \n",
       "3                    244  ...               2             0               1   \n",
       "4                     64  ...               7             0               0   \n",
       "\n",
       "   title_pos_verb  text_pos_adj text_pos_adv text_pos_noun  text_pos_num  \\\n",
       "id                                                                         \n",
       "0               1           102           16           234             3   \n",
       "1               1            81           13           208             3   \n",
       "2               2           155           33           385             4   \n",
       "3               1            74            5           158             6   \n",
       "4               1            13            5            54             2   \n",
       "\n",
       "    text_pos_pron  text_pos_verb  \n",
       "id                                \n",
       "0               0             68  \n",
       "1               3             45  \n",
       "2               3             92  \n",
       "3              12             44  \n",
       "4               0             11  \n",
       "\n",
       "[5 rows x 29 columns]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# View data\n",
    "news_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b3302fa4",
   "metadata": {},
   "source": [
    "# Model training"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dd646486",
   "metadata": {},
   "source": [
    "## Attempt 1: Decision Tree"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1177c07d",
   "metadata": {},
   "source": [
    "### Finding out what variables are the most impactful"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "7a8e0be5",
   "metadata": {},
   "outputs": [],
   "source": [
    "def auto_kfold_decision_tree(variables):\n",
    "    # Init list to record scores\n",
    "    scores = {}\n",
    "    \n",
    "    # Response\n",
    "    news_res = news_data['label_translated']\n",
    "    \n",
    "    # Train for every variable\n",
    "    for variable in variables:\n",
    "        if variable.dtype == np.dtype('int64'):\n",
    "            # Numeric variable\n",
    "            news_cat = pd.DataFrame(variable)\n",
    "            newsdata_ohe = pd.concat([news_cat, news_res], sort = False, axis = 1).reindex(index=news_cat.index)\n",
    "        elif variable.dtype == np.dtype('object'):\n",
    "            # Categorical variables\n",
    "            # Import the encoder from sklearn\n",
    "            ohe = OneHotEncoder()\n",
    "            # OneHotEncoding of categorical predictors (not the response)\n",
    "            news_cat = pd.DataFrame(variable)\n",
    "\n",
    "            ohe.fit(news_cat)\n",
    "            news_cat_ohe = pd.DataFrame(ohe.transform(news_cat).toarray(), columns=ohe.get_feature_names(news_cat.columns))\n",
    "            \n",
    "            newsdata_ohe = pd.concat([news_cat_ohe, news_res], sort = False, axis = 1).reindex(index=news_cat_ohe.index)\n",
    "\n",
    "        # Extract Response and Predictors\n",
    "        y = pd.DataFrame(newsdata_ohe['label_translated'])\n",
    "        X = pd.DataFrame(newsdata_ohe.drop('label_translated', axis = 1))\n",
    "\n",
    "        # Split the Dataset into Train and Test\n",
    "        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.3)\n",
    "\n",
    "        # Decision Tree using Train Data\n",
    "        dectree = DecisionTreeClassifier(max_depth = 4)  # change max_depth to experiment\n",
    "\n",
    "        skf = StratifiedKFold(n_splits=5, shuffle=True, random_state=17)\n",
    "        cv_results = cross_val_score(dectree, X_train, y_train['label_translated'], cv=skf)\n",
    "\n",
    "        # Record score\n",
    "        scores[variable.name] = cv_results\n",
    "        \n",
    "    return scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "ae38a7de",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictor: title_wordcount, score: 0.70\n",
      "Predictor: title_charcount, score: 0.68\n",
      "Predictor: title_pos_adj, score: 0.67\n",
      "Predictor: stopwords_count_text, score: 0.64\n",
      "Predictor: text_pos_noun, score: 0.63\n",
      "Predictor: text_charcount, score: 0.63\n",
      "Predictor: title_pos_noun, score: 0.63\n",
      "Predictor: text_wordcount, score: 0.63\n",
      "Predictor: text_pos_adj, score: 0.63\n",
      "Predictor: text_pos_verb, score: 0.63\n",
      "Predictor: text_pos_num, score: 0.62\n",
      "Predictor: text_pos_adv, score: 0.62\n",
      "Predictor: stopwords_count_title, score: 0.58\n",
      "Predictor: title_pos_verb, score: 0.57\n",
      "Predictor: text_pos_pron, score: 0.55\n",
      "Predictor: title_polarity, score: 0.54\n",
      "Predictor: text_emotion, score: 0.54\n",
      "Predictor: title_emotion, score: 0.54\n",
      "Predictor: title_pos_pron, score: 0.52\n",
      "Predictor: title_pos_adv, score: 0.51\n",
      "Predictor: text_polarity, score: 0.51\n",
      "Predictor: title_pos_num, score: 0.50\n"
     ]
    }
   ],
   "source": [
    "variables = [news_data['title_charcount'], \n",
    "             news_data['text_charcount'], \n",
    "             news_data['title_wordcount'], \n",
    "             news_data['text_wordcount'], \n",
    "             news_data['stopwords_count_text'], \n",
    "             news_data['stopwords_count_title'], \n",
    "             news_data['title_polarity'], \n",
    "             news_data['text_polarity'], \n",
    "             news_data['title_emotion'], \n",
    "             news_data['text_emotion'], \n",
    "             news_data['title_pos_adj'], \n",
    "             news_data['title_pos_adv'], \n",
    "             news_data['title_pos_noun'], \n",
    "             news_data['title_pos_num'], \n",
    "             news_data['title_pos_pron'], \n",
    "             news_data['title_pos_verb'], \n",
    "             news_data['text_pos_adj'], \n",
    "             news_data['text_pos_adv'], \n",
    "             news_data['text_pos_noun'], \n",
    "             news_data['text_pos_num'], \n",
    "             news_data['text_pos_pron'], \n",
    "             news_data['text_pos_verb']\n",
    "            ]\n",
    "\n",
    "scores = auto_kfold_decision_tree(variables)\n",
    "\n",
    "\n",
    "for key, value in sorted(scores.items(), key=lambda x: x[1].mean(), reverse=True):\n",
    "    print(f\"Predictor: {key}, score: {value.mean():.2f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4ef089c0",
   "metadata": {},
   "source": [
    "### Training with top predictors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "2f2b2dde",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 20584 entries, 0 to 20583\n",
      "Data columns (total 6 columns):\n",
      " #   Column                Non-Null Count  Dtype \n",
      "---  ------                --------------  ----- \n",
      " 0   title_wordcount       20584 non-null  int64 \n",
      " 1   title_pos_adj         20584 non-null  int64 \n",
      " 2   stopwords_count_text  20584 non-null  int64 \n",
      " 3   title_pos_noun        20584 non-null  int64 \n",
      " 4   text_wordcount        20584 non-null  int64 \n",
      " 5   label_translated      20584 non-null  object\n",
      "dtypes: int64(5), object(1)\n",
      "memory usage: 1.1+ MB\n"
     ]
    }
   ],
   "source": [
    "news_num = news_data[['title_wordcount','title_pos_adj','stopwords_count_text','title_pos_noun','text_wordcount']]\n",
    "news_res = news_data['label_translated']\n",
    "news_final = pd.concat([news_num,news_res],sort=False,axis=1).reindex(index=news_num.index)\n",
    "# Check the final dataframe\n",
    "news_final.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "01a797d6",
   "metadata": {},
   "source": [
    "### Create and fit model using decision tree\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "65768c7c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extract Response and Predictors\n",
    "y = pd.DataFrame(news_final['label_translated'])\n",
    "X = pd.DataFrame(news_final.drop('label_translated', axis = 1))\n",
    "\n",
    "# Split the Dataset into Train and Test\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.3)\n",
    "\n",
    "# Decision Tree using Train Data\n",
    "dectree = DecisionTreeClassifier(max_depth = 4)  # change max_depth to experiment\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52285f78",
   "metadata": {},
   "source": [
    "### Stratified k- fold for decision tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "1b6f3aaa",
   "metadata": {},
   "outputs": [],
   "source": [
    "skf = StratifiedKFold(n_splits=5, shuffle=True, random_state=17)\n",
    "cv_results = cross_val_score(dectree, X_train, y_train['label_translated'], cv=skf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "0a761df4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0.72935461, 0.73560028, 0.7369882 , 0.72405415, 0.73030198]),\n",
       " 0.731259844283577)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results, cv_results.mean()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a435d6f3",
   "metadata": {},
   "source": [
    "### Fitting data into decision tree\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "7d698952",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeClassifier(max_depth=4)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dectree.fit(X_train, y_train)                    # train the decision tree model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c9e19b8d",
   "metadata": {},
   "source": [
    "### Decision tree evaluation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9a2344ec",
   "metadata": {},
   "source": [
    "#### Plotting decision tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "e2ef593c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(669.6, 1141.5600000000002, 'title_wordcount <= 11.5\\ngini = 0.5\\nsamples = 14408\\nvalue = [7111, 7297]\\nclass = Not Fake'),\n",
       " Text(334.8, 815.4000000000001, 'title_wordcount <= 7.5\\ngini = 0.4\\nsamples = 5889\\nvalue = [4262, 1627]\\nclass = Fake'),\n",
       " Text(167.4, 489.24, 'title_pos_noun <= 4.5\\ngini = 0.18\\nsamples = 1572\\nvalue = [1415, 157]\\nclass = Fake'),\n",
       " Text(83.7, 163.07999999999993, '\\n  (...)  \\n'),\n",
       " Text(251.10000000000002, 163.07999999999993, '\\n  (...)  \\n'),\n",
       " Text(502.20000000000005, 489.24, 'stopwords_count_text <= 46.5\\ngini = 0.449\\nsamples = 4317\\nvalue = [2847, 1470]\\nclass = Fake'),\n",
       " Text(418.5, 163.07999999999993, '\\n  (...)  \\n'),\n",
       " Text(585.9, 163.07999999999993, '\\n  (...)  \\n'),\n",
       " Text(1004.4000000000001, 815.4000000000001, 'stopwords_count_text <= 230.5\\ngini = 0.445\\nsamples = 8519\\nvalue = [2849, 5670]\\nclass = Not Fake'),\n",
       " Text(837.0, 489.24, 'stopwords_count_text <= 47.5\\ngini = 0.498\\nsamples = 3202\\nvalue = [1708, 1494]\\nclass = Fake'),\n",
       " Text(753.3000000000001, 163.07999999999993, '\\n  (...)  \\n'),\n",
       " Text(920.7, 163.07999999999993, '\\n  (...)  \\n'),\n",
       " Text(1171.8, 489.24, 'title_pos_adj <= 1.5\\ngini = 0.337\\nsamples = 5317\\nvalue = [1141, 4176]\\nclass = Not Fake'),\n",
       " Text(1088.1000000000001, 163.07999999999993, '\\n  (...)  \\n'),\n",
       " Text(1255.5, 163.07999999999993, '\\n  (...)  \\n')]"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1728x1728 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the trained Decision Tree\n",
    "f = plt.figure(figsize=(24,24))\n",
    "plot_tree(dectree, filled=True, rounded=True, max_depth=2,\n",
    "          feature_names=X_train.columns, \n",
    "          class_names=[\"Fake\",\"Not Fake\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b13d1ad0",
   "metadata": {},
   "source": [
    "#### Confusion matrix\n",
    "\n",
    "Print the Classification Accuracy and all other Accuracy Measures from the Confusion Matrix.  \n",
    "\n",
    "| Confusion Matrix  |       |        |        |      \n",
    "| :---              | :---: | :----: | :----: |         \n",
    "| Actual Negative   |  (0)  |   TN   |   FP   |             \n",
    "| Actual Positive   |  (1)  |   FN   |   TP   |       \n",
    "|                   |       |   (0)   |   (1)   |       \n",
    "|                   |       | Predicted Negative    |   Predicted Postitive  |     \n",
    "\n",
    "\n",
    "* `TPR = TP / (TP + FN)` : True Positive Rate = True Positives / All Positives    \n",
    "* `TNR = TN / (TN + FP)` : True Negative Rate = True Negatives / All Negatives    \n",
    "\n",
    "* `FPR = FP / (TN + FP)` : False Positive Rate = False Positives / All Negatives \n",
    "* `FNR = FN / (TP + FN)` : False Negative Rate = False Negatives / All Positives "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "f0fbd38a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Data\n",
      "Accuracy  :\t 0.7318156579677957\n",
      "\n",
      "TPR Train :\t 0.6971358092366726\n",
      "TNR Train :\t 0.7674026156658698\n",
      "\n",
      "FPR Train :\t 0.2325973843341302\n",
      "FNR Train :\t 0.3028641907633274\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Predict the Response corresponding to Predictors\n",
    "y_train_pred = dectree.predict(X_train)\n",
    "\n",
    "# Print the Classification Accuracy\n",
    "print(\"Train Data\")\n",
    "print(\"Accuracy  :\\t\", dectree.score(X_train, y_train))\n",
    "print()\n",
    "\n",
    "# Print the Accuracy Measures from the Confusion Matrix\n",
    "cmTrain = confusion_matrix(y_train, y_train_pred)\n",
    "tpTrain = cmTrain[1][1] # True Positives : Good (1) predicted Good (1)\n",
    "fpTrain = cmTrain[0][1] # False Positives : Bad (0) predicted Good (1)\n",
    "tnTrain = cmTrain[0][0] # True Negatives : Bad (0) predicted Bad (0)\n",
    "fnTrain = cmTrain[1][0] # False Negatives : Good (1) predicted Bad (0)\n",
    "\n",
    "print(\"TPR Train :\\t\", (tpTrain/(tpTrain + fnTrain)))\n",
    "print(\"TNR Train :\\t\", (tnTrain/(tnTrain + fpTrain)))\n",
    "print()\n",
    "\n",
    "print(\"FPR Train :\\t\", (fpTrain/(tnTrain + fpTrain)))\n",
    "print(\"FNR Train :\\t\", (fnTrain/(tpTrain + fnTrain)))\n",
    "\n",
    "# Plot the two-way Confusion Matrix\n",
    "sb.heatmap(confusion_matrix(y_train, y_train_pred), \n",
    "           annot = True, fmt=\".0f\", annot_kws={\"size\": 18})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "994ba766",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Data\n",
      "Accuracy  :\t 0.7313795336787565\n",
      "\n",
      "TPR Test :\t 0.6869537067011978\n",
      "TNR Test :\t 0.7758341431810819\n",
      "\n",
      "FPR Test :\t 0.22416585681891804\n",
      "FNR Test :\t 0.3130462932988022\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Predict the Response corresponding to Predictors\n",
    "y_test_pred = dectree.predict(X_test)\n",
    "\n",
    "# Print the Classification Accuracy\n",
    "print(\"Test Data\")\n",
    "print(\"Accuracy  :\\t\", dectree.score(X_test, y_test))\n",
    "print()\n",
    "\n",
    "# Print the Accuracy Measures from the Confusion Matrix\n",
    "cmTest = confusion_matrix(y_test, y_test_pred)\n",
    "tpTest = cmTest[1][1] # True Positives : Good (1) predicted Good (1)\n",
    "fpTest = cmTest[0][1] # False Positives : Bad (0) predicted Good (1)\n",
    "tnTest = cmTest[0][0] # True Negatives : Bad (0) predicted Bad (0)\n",
    "fnTest = cmTest[1][0] # False Negatives : Good (1) predicted Bad (0)\n",
    "\n",
    "print(\"TPR Test :\\t\", (tpTest/(tpTest + fnTest)))\n",
    "print(\"TNR Test :\\t\", (tnTest/(tnTest + fpTest)))\n",
    "print()\n",
    "\n",
    "print(\"FPR Test :\\t\", (fpTest/(fpTest + tnTest)))\n",
    "print(\"FNR Test :\\t\", (fnTest/(fnTest + tpTest)))\n",
    "\n",
    "# Plot the two-way Confusion Matrix\n",
    "sb.heatmap(confusion_matrix(y_test, y_test_pred), \n",
    "           annot = True, fmt=\".0f\", annot_kws={\"size\": 18})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "144b538d",
   "metadata": {},
   "source": [
    "## Attempt 2: Random forest classifier\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "80ec84d2",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extract Response and Predictors\n",
    "y = pd.DataFrame(news_final['label_translated'])\n",
    "X = pd.DataFrame(news_final.drop('label_translated', axis = 1))\n",
    "\n",
    "# Split the Dataset into Train and Test\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "5497df12",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create the Random Forest object\n",
    "rforest = RandomForestClassifier(n_estimators = 100,  # n_estimators denote number of trees\n",
    "                                 max_depth = 4)       # set the maximum depth of each tree\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "46ccfb18",
   "metadata": {},
   "source": [
    "### Stratified k-fold verification for random forest classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "02ff4e79",
   "metadata": {},
   "outputs": [],
   "source": [
    "skf = StratifiedKFold(n_splits=5, shuffle=True, random_state=17)\n",
    "cv_results = cross_val_score(rforest, X_train, y_train['label_translated'], cv=skf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "91ab9797",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0.73004858, 0.73004858, 0.74739764, 0.74800417, 0.72370705]),\n",
       " 0.7358412013331981)"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results, cv_results.mean()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "16c4e254",
   "metadata": {},
   "source": [
    "### Fitting data into forest "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "5e6534b6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(max_depth=4)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Fit Random Forest on Train Data\n",
    "rforest.fit(X_train, y_train.label_translated.ravel())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d03112ae",
   "metadata": {},
   "source": [
    "### Random Forest model evaluation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f071508c",
   "metadata": {},
   "source": [
    "### Confusion Matrix\n",
    "\n",
    "Print the Classification Accuracy and all other Accuracy Measures from the Confusion Matrix.  \n",
    "\n",
    "| Confusion Matrix  |       |        |        |      \n",
    "| :---              | :---: | :----: | :----: |         \n",
    "| Actual Negative   |  (0)  |   TN   |   FP   |             \n",
    "| Actual Positive   |  (1)  |   FN   |   TP   |       \n",
    "|                   |       |   (0)   |   (1)   |       \n",
    "|                   |       | Predicted Negative    |   Predicted Postitive  |     \n",
    "\n",
    "\n",
    "* `TPR = TP / (TP + FN)` : True Positive Rate = True Positives / All Positives    \n",
    "* `TNR = TN / (TN + FP)` : True Negative Rate = True Negatives / All Negatives    \n",
    "\n",
    "* `FPR = FP / (TN + FP)` : False Positive Rate = False Positives / All Negatives \n",
    "* `FNR = FN / (TP + FN)` : False Negative Rate = False Negatives / All Positives "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "60a3893c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Data\n",
      "Accuracy  :\t 0.7384092171016102\n",
      "\n",
      "TPR Train :\t 0.7238424326192122\n",
      "TNR Train :\t 0.753101909940053\n",
      "\n",
      "FPR Train :\t 0.24689809005994703\n",
      "FNR Train :\t 0.27615756738078784\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Predict the Response corresponding to Predictors\n",
    "y_train_pred = rforest.predict(X_train)\n",
    "\n",
    "# Print the Classification Accuracy\n",
    "print(\"Train Data\")\n",
    "print(\"Accuracy  :\\t\", rforest.score(X_train, y_train))\n",
    "print()\n",
    "\n",
    "# Print the Accuracy Measures from the Confusion Matrix\n",
    "cmTrain = confusion_matrix(y_train, y_train_pred)\n",
    "tpTrain = cmTrain[1][1] # True Positives : Good (1) predicted Good (1)\n",
    "fpTrain = cmTrain[0][1] # False Positives : Bad (0) predicted Good (1)\n",
    "tnTrain = cmTrain[0][0] # True Negatives : Bad (0) predicted Bad (0)\n",
    "fnTrain = cmTrain[1][0] # False Negatives : Good (1) predicted Bad (0)\n",
    "\n",
    "print(\"TPR Train :\\t\", (tpTrain/(tpTrain + fnTrain)))\n",
    "print(\"TNR Train :\\t\", (tnTrain/(tnTrain + fpTrain)))\n",
    "print()\n",
    "\n",
    "print(\"FPR Train :\\t\", (fpTrain/(tnTrain + fpTrain)))\n",
    "print(\"FNR Train :\\t\", (fnTrain/(tpTrain + fnTrain)))\n",
    "\n",
    "# Plot the two-way Confusion Matrix\n",
    "sb.heatmap(confusion_matrix(y_train, y_train_pred), \n",
    "           annot = True, fmt=\".0f\", annot_kws={\"size\": 18})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "4cfe20ac",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Data\n",
      "Accuracy  :\t 0.7291126943005182\n",
      "\n",
      "TPR Test :\t 0.7280228498889242\n",
      "TNR Test :\t 0.7302479338842975\n",
      "\n",
      "FPR Test :\t 0.26975206611570246\n",
      "FNR Test :\t 0.27197715011107587\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Predict the Response corresponding to Predictors\n",
    "y_test_pred = rforest.predict(X_test)\n",
    "\n",
    "# Print the Classification Accuracy\n",
    "print(\"Test Data\")\n",
    "print(\"Accuracy  :\\t\", rforest.score(X_test, y_test))\n",
    "print()\n",
    "\n",
    "# Print the Accuracy Measures from the Confusion Matrix\n",
    "cmTest = confusion_matrix(y_test, y_test_pred)\n",
    "tpTest = cmTest[1][1] # True Positives : Good (1) predicted Good (1)\n",
    "fpTest = cmTest[0][1] # False Positives : Bad (0) predicted Good (1)\n",
    "tnTest = cmTest[0][0] # True Negatives : Bad (0) predicted Bad (0)\n",
    "fnTest = cmTest[1][0] # False Negatives : Good (1) predicted Bad (0)\n",
    "\n",
    "print(\"TPR Test :\\t\", (tpTest/(tpTest + fnTest)))\n",
    "print(\"TNR Test :\\t\", (tnTest/(tnTest + fpTest)))\n",
    "print()\n",
    "\n",
    "print(\"FPR Test :\\t\", (fpTest/(fpTest + tnTest)))\n",
    "print(\"FNR Test :\\t\", (fnTest/(fnTest + tpTest)))\n",
    "\n",
    "# Plot the two-way Confusion Matrix\n",
    "sb.heatmap(confusion_matrix(y_test, y_test_pred), \n",
    "           annot = True, fmt=\".0f\", annot_kws={\"size\": 18})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "029443e1",
   "metadata": {},
   "source": [
    "### Trying to improve accuracy by hyperparameter tuning using Cross-Validation(CV)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "bf5dabf8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extract Response and Predictors\n",
    "y = pd.DataFrame(news_final['label_translated'])\n",
    "X = pd.DataFrame(news_final.drop('label_translated', axis = 1))\n",
    "\n",
    "# Split the Dataset into Train and Test\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "64be3c4a",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, estimator=RandomForestClassifier(),\n",
       "             param_grid={'max_depth': array([ 2,  3,  4,  5,  6,  7,  8,  9, 10]),\n",
       "                         'n_estimators': array([ 100,  200,  300,  400,  500,  600,  700,  800,  900, 1000])},\n",
       "             scoring='accuracy')"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Define the Hyper-parameter Grid to search on, in case of Random Forest\n",
    "param_grid = {'n_estimators': np.arange(100,1001,100),   # number of trees 100, 200, ..., 1000\n",
    "              'max_depth': np.arange(2, 11)}             # depth of trees 2, 3, 4, 5, ..., 10\n",
    "\n",
    "# Create the Hyper-parameter Grid\n",
    "hpGrid = GridSearchCV(RandomForestClassifier(),   # the model family\n",
    "                      param_grid,                 # the search grid\n",
    "                      cv = 5,                     # 5-fold cross-validation\n",
    "                      scoring = 'accuracy')       # score to evaluate\n",
    "\n",
    "# Train the models using Cross-Validation\n",
    "hpGrid.fit(X_train, y_train.label_translated.ravel())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "85cba478",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RandomForestClassifier(max_depth=10, n_estimators=500)\n",
      "0.7503467765187747\n"
     ]
    }
   ],
   "source": [
    "# Fetch the best Model or the best set of Hyper-parameters\n",
    "print(hpGrid.best_estimator_)\n",
    "\n",
    "# Print the score (accuracy) of the best Model after CV\n",
    "print(np.abs(hpGrid.best_score_))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "68e762e4",
   "metadata": {},
   "source": [
    "#### Best result is max depth = 10 n_estimators = 1000 accuracy = 0.76\n",
    "#### This prompts us to use another machine learning algorithm to try to improve our accuracy"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}