ML_Random_Forest.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "bd8adac1-6405-435d-a245-0289ce71152f",
   "metadata": {},
   "source": [
    "### Importing necessary library"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "6c690210-bc4c-49ff-a2c7-c915e37e88c4",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier, export_graphviz\n",
    "from sklearn import metrics\n",
    "\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import accuracy_score, classification_report, confusion_matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d3543198-e6ba-4056-95a5-a4ae5b4938e7",
   "metadata": {},
   "source": [
    "### Load the CSV file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "93edc606-8de8-4cc9-9fc7-7a187fb7bce7",
   "metadata": {},
   "outputs": [],
   "source": [
    "CSV_DataFrame = pd.read_csv(\"adult.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "4bf66c9a-7c6f-4fd2-a657-ced7bfba897c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Classes in the data set\n",
    "classes = [\"<=50K\", \">50K\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1aca2a10-5e0b-4e0a-8d8f-43a3ad3efbb6",
   "metadata": {},
   "source": [
    "### Encoding the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "91c38ca0-4f9e-4e98-a669-91144bc6e714",
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Initialize LabelEncoder\n",
    "# label_encoder = LabelEncoder()\n",
    "\n",
    "# categorical_columns = [\"work_class\", \"education\", \"marital_status\", \"occupation\", \"relationship\", \"race\", \"sex\", \"native_country\"]\n",
    "\n",
    "# # Iterate over each categorical column and encode its values\n",
    "# for column in categorical_columns:\n",
    "#     CSV_DataFrame[column] = label_encoder.fit_transform(CSV_DataFrame[column])\n",
    "#     print(f\"Encoded labels for column '{column}': {label_encoder.classes_}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "faa8abc0-a09d-4c02-b6eb-315baf681c07",
   "metadata": {},
   "source": [
    "### Feeding the data into the model to train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "d0fe3a12-4239-49d2-b393-e7331978f14f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Separate features and target variable\n",
    "X = CSV_DataFrame.drop(columns=['income'])\n",
    "y = CSV_DataFrame['income']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "3def9475-6bb3-4add-b800-48d3c834510f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split the data into training and testing sets\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=69)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d7f6eef0-e562-49c3-8a30-8e861e3b7f85",
   "metadata": {},
   "source": [
    "### With no over-sampled data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "9035e7a5-e56b-4f00-9e27-2e5d6e3c5918",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "income\n",
      "0    7398\n",
      "1    2256\n",
      "Name: count, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "income_counts = y_test.value_counts()\n",
    "print(income_counts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "abeae532-8989-45b8-a931-ff2c53bcb538",
   "metadata": {},
   "outputs": [],
   "source": [
    "random_forest_classifier = RandomForestClassifier(random_state=69)\n",
    "\n",
    "# Train the classifier on the resampled training data\n",
    "random_forest_classifier.fit(X_train, y_train)\n",
    "\n",
    "# prediction\n",
    "y_pred = random_forest_classifier.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "56626a11-da18-4b91-b2a8-a06db9ee9c51",
   "metadata": {},
   "source": [
    "#### Evaluation Matrix and Classification Report no oversampled model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "3177824d-0f85-4e03-b461-48fac2d12aac",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8467992541951522\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model\n",
    "accuracy = accuracy_score(y_test, y_pred)\n",
    "print(\"Accuracy:\", accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "e1b98040-ad6d-4d66-b819-c90ac24b160e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report:\n"
     ]
    },
    {
     "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>&lt;=50K</th>\n",
       "      <th>&gt;50K</th>\n",
       "      <th>accuracy</th>\n",
       "      <th>macro avg</th>\n",
       "      <th>weighted avg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>precision</th>\n",
       "      <td>0.886307</td>\n",
       "      <td>0.694932</td>\n",
       "      <td>0.846799</td>\n",
       "      <td>0.790620</td>\n",
       "      <td>0.841586</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.917816</td>\n",
       "      <td>0.613918</td>\n",
       "      <td>0.846799</td>\n",
       "      <td>0.765867</td>\n",
       "      <td>0.846799</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.901786</td>\n",
       "      <td>0.651918</td>\n",
       "      <td>0.846799</td>\n",
       "      <td>0.776852</td>\n",
       "      <td>0.843396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>7398.000000</td>\n",
       "      <td>2256.000000</td>\n",
       "      <td>0.846799</td>\n",
       "      <td>9654.000000</td>\n",
       "      <td>9654.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 <=50K         >50K  accuracy    macro avg  weighted avg\n",
       "precision     0.886307     0.694932  0.846799     0.790620      0.841586\n",
       "recall        0.917816     0.613918  0.846799     0.765867      0.846799\n",
       "f1-score      0.901786     0.651918  0.846799     0.776852      0.843396\n",
       "support    7398.000000  2256.000000  0.846799  9654.000000   9654.000000"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Print classification report\n",
    "print(\"Classification Report:\")\n",
    "\n",
    "actual_classes = [classes[x] for x in y_test]\n",
    "predicted_classes = [classes[x] for x in y_pred]\n",
    "\n",
    "report_dict = classification_report(actual_classes, predicted_classes, output_dict=True)\n",
    "classification_df = pd.DataFrame(report_dict)\n",
    "classification_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "c13bfc74-ffd9-455d-bf33-67b31b9863f2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 800x600 with 0 Axes>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 800x600 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Generate confusion matrix\n",
    "print(\"Confusion Matrix:\")\n",
    "\n",
    "conf_matrix = confusion_matrix(y_test, y_pred)\n",
    "sns.heatmap(conf_matrix, annot=True, fmt=\"d\", cmap=\"Blues\", xticklabels=classes, yticklabels=classes)\n",
    "\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.figure(figsize = (8,6))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a72252e1-1a1f-4ca0-9132-c5ad592f8d11",
   "metadata": {},
   "source": [
    "#### With over-sampling of data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "1ad740b2-51b9-43ad-8c56-2a0dd9fcea2f",
   "metadata": {},
   "outputs": [],
   "source": [
    "from imblearn.over_sampling import SMOTE\n",
    "\n",
    "# Over sampling applying only for the training dataset\n",
    "# Apply SMOTE for oversampling\n",
    "smote = SMOTE(random_state=69)\n",
    "X_train_resampled, y_train_resampled = smote.fit_resample(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "961a02c9-73f0-41f6-8464-cb4467fd6de3",
   "metadata": {},
   "outputs": [],
   "source": [
    "random_forest_classifier_oversample = RandomForestClassifier(random_state=69)\n",
    "\n",
    "# Train the classifier on the resampled training data\n",
    "random_forest_classifier_oversample.fit(X_train_resampled, y_train_resampled)\n",
    "\n",
    "# prediction\n",
    "y_pred = random_forest_classifier_oversample.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "77d3b0a8-de39-40d9-8646-686ddb2cea34",
   "metadata": {},
   "source": [
    "#### Evaluation Matrix and Classification Report over-sampled model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "41b67bfe-856f-455e-a506-c4f8f94d1fb8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8395483737310959\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model\n",
    "accuracy = accuracy_score(y_test, y_pred)\n",
    "print(\"Accuracy:\", accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "fbdbdaae-7659-4b43-bda6-de197d3ac041",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report:\n"
     ]
    },
    {
     "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>&lt;=50K</th>\n",
       "      <th>&gt;50K</th>\n",
       "      <th>accuracy</th>\n",
       "      <th>macro avg</th>\n",
       "      <th>weighted avg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>precision</th>\n",
       "      <td>0.893872</td>\n",
       "      <td>0.658591</td>\n",
       "      <td>0.839548</td>\n",
       "      <td>0.776232</td>\n",
       "      <td>0.838890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.897134</td>\n",
       "      <td>0.650709</td>\n",
       "      <td>0.839548</td>\n",
       "      <td>0.773922</td>\n",
       "      <td>0.839548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.895500</td>\n",
       "      <td>0.654627</td>\n",
       "      <td>0.839548</td>\n",
       "      <td>0.775063</td>\n",
       "      <td>0.839212</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>7398.000000</td>\n",
       "      <td>2256.000000</td>\n",
       "      <td>0.839548</td>\n",
       "      <td>9654.000000</td>\n",
       "      <td>9654.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 <=50K         >50K  accuracy    macro avg  weighted avg\n",
       "precision     0.893872     0.658591  0.839548     0.776232      0.838890\n",
       "recall        0.897134     0.650709  0.839548     0.773922      0.839548\n",
       "f1-score      0.895500     0.654627  0.839548     0.775063      0.839212\n",
       "support    7398.000000  2256.000000  0.839548  9654.000000   9654.000000"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Print classification report\n",
    "print(\"Classification Report:\")\n",
    "\n",
    "actual_classes = [classes[x] for x in y_test]\n",
    "predicted_classes = [classes[x] for x in y_pred]\n",
    "\n",
    "report_dict = classification_report(actual_classes, predicted_classes, output_dict=True)\n",
    "classification_df = pd.DataFrame(report_dict)\n",
    "classification_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "0148381e-4b31-4815-b9b1-92f321db5578",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 800x600 with 0 Axes>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 800x600 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Generate confusion matrix\n",
    "print(\"Confusion Matrix:\")\n",
    "\n",
    "conf_matrix = confusion_matrix(y_test, y_pred)\n",
    "sns.heatmap(conf_matrix, annot=True, fmt=\"d\", cmap=\"Blues\", xticklabels=classes, yticklabels=classes)\n",
    "\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.figure(figsize = (8,6))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6acf9004-699a-475b-8314-632e896c9d33",
   "metadata": {},
   "source": [
    "#### Saving the random forest model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "4a508372-cd72-48eb-9066-81e6ee97cc32",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['random_forest_model.pkl']"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from joblib import dump\n",
    "\n",
    "# Save Random Forest model\n",
    "dump(random_forest_classifier, 'random_forest_model.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8749b4c2-0e16-42e4-ad10-bb216fcf6ce7",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.9.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}