Music-Genre-Classifier.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "6972400f",
   "metadata": {},
   "source": [
    "# CS4661 Project Report: Music Genre Classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "f22b471c",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn import metrics\n",
    "import math\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "af0f46bf",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# \"read_csv\" is a pandas function that we use here to read the data file \"train (1).csv\"\n",
    "train_df = pd.read_csv(\"shai-music-genre-classification//train (1).csv\")\n",
    "\n",
    "# Create a variable for the mean of values in the \"instrumentalness\" column\n",
    "X_mean_instrum = train_df['instrumentalness'].mean()\n",
    "\n",
    "# Create a variable for the mean of values in the \"Popularity\" column\n",
    "X_mean_popul = train_df['Popularity'].mean()\n",
    "\n",
    "# Create a variable for the mean of values in the \"key\" column\n",
    "X_mean_key = train_df['key'].mean()\n",
    "\n",
    "# Replace NaNs in these columns with the mean of values for their respective columns\n",
    "train_df['instrumentalness'].fillna(value=X_mean_instrum, inplace=True)\n",
    "train_df['Popularity'].fillna(value=X_mean_popul, inplace=True)\n",
    "train_df['key'].fillna(value=X_mean_key, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "e014c127",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Since label value 10 corresponds to 'rock', we will replaces all 10 values with 1 and all others with 0 in the 'Class' \n",
    "# column so that we can have a binary classification\n",
    "not_rock = train_df['Class'] != 10\n",
    "is_rock = train_df['Class'] == 10\n",
    "\n",
    "train_df.loc[not_rock, 'Class'] = 0\n",
    "train_df.loc[is_rock, 'Class'] = 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c201b334",
   "metadata": {},
   "source": [
    "# Testing Accuracy and AUC when NaN values are replaced with the mean of all other values in that column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "2a495929",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(12236, 14)\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>Popularity</th>\n",
       "      <th>danceability</th>\n",
       "      <th>energy</th>\n",
       "      <th>key</th>\n",
       "      <th>loudness</th>\n",
       "      <th>mode</th>\n",
       "      <th>speechiness</th>\n",
       "      <th>acousticness</th>\n",
       "      <th>instrumentalness</th>\n",
       "      <th>liveness</th>\n",
       "      <th>valence</th>\n",
       "      <th>tempo</th>\n",
       "      <th>duration_in min/ms</th>\n",
       "      <th>time_signature</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>9188</th>\n",
       "      <td>49.0</td>\n",
       "      <td>0.184</td>\n",
       "      <td>0.207</td>\n",
       "      <td>2.0</td>\n",
       "      <td>-17.119</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0330</td>\n",
       "      <td>0.940000</td>\n",
       "      <td>0.931000</td>\n",
       "      <td>0.0772</td>\n",
       "      <td>0.089</td>\n",
       "      <td>79.979</td>\n",
       "      <td>251173.000000</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7086</th>\n",
       "      <td>69.0</td>\n",
       "      <td>0.553</td>\n",
       "      <td>0.872</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-2.832</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0782</td>\n",
       "      <td>0.198000</td>\n",
       "      <td>0.178129</td>\n",
       "      <td>0.0659</td>\n",
       "      <td>0.638</td>\n",
       "      <td>172.138</td>\n",
       "      <td>3.676217</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3463</th>\n",
       "      <td>22.0</td>\n",
       "      <td>0.118</td>\n",
       "      <td>0.964</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-4.060</td>\n",
       "      <td>1</td>\n",
       "      <td>0.4360</td>\n",
       "      <td>0.000120</td>\n",
       "      <td>0.139000</td>\n",
       "      <td>0.3370</td>\n",
       "      <td>0.107</td>\n",
       "      <td>200.003</td>\n",
       "      <td>269787.000000</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1897</th>\n",
       "      <td>58.0</td>\n",
       "      <td>0.634</td>\n",
       "      <td>0.648</td>\n",
       "      <td>3.0</td>\n",
       "      <td>-6.781</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0412</td>\n",
       "      <td>0.361000</td>\n",
       "      <td>0.178129</td>\n",
       "      <td>0.6190</td>\n",
       "      <td>0.667</td>\n",
       "      <td>95.989</td>\n",
       "      <td>4.130200</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1907</th>\n",
       "      <td>32.0</td>\n",
       "      <td>0.413</td>\n",
       "      <td>0.690</td>\n",
       "      <td>10.0</td>\n",
       "      <td>-6.868</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0299</td>\n",
       "      <td>0.212000</td>\n",
       "      <td>0.024100</td>\n",
       "      <td>0.3140</td>\n",
       "      <td>0.483</td>\n",
       "      <td>95.260</td>\n",
       "      <td>267533.000000</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9160</th>\n",
       "      <td>37.0</td>\n",
       "      <td>0.479</td>\n",
       "      <td>0.909</td>\n",
       "      <td>2.0</td>\n",
       "      <td>-7.404</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0716</td>\n",
       "      <td>0.000091</td>\n",
       "      <td>0.758000</td>\n",
       "      <td>0.6340</td>\n",
       "      <td>0.278</td>\n",
       "      <td>111.892</td>\n",
       "      <td>310467.000000</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9859</th>\n",
       "      <td>71.0</td>\n",
       "      <td>0.671</td>\n",
       "      <td>0.256</td>\n",
       "      <td>7.0</td>\n",
       "      <td>-10.743</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0387</td>\n",
       "      <td>0.608000</td>\n",
       "      <td>0.178129</td>\n",
       "      <td>0.0737</td>\n",
       "      <td>0.694</td>\n",
       "      <td>115.871</td>\n",
       "      <td>233560.000000</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11513</th>\n",
       "      <td>39.0</td>\n",
       "      <td>0.579</td>\n",
       "      <td>0.636</td>\n",
       "      <td>2.0</td>\n",
       "      <td>-8.574</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0276</td>\n",
       "      <td>0.082100</td>\n",
       "      <td>0.922000</td>\n",
       "      <td>0.4460</td>\n",
       "      <td>0.898</td>\n",
       "      <td>132.652</td>\n",
       "      <td>280853.000000</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1688</th>\n",
       "      <td>66.0</td>\n",
       "      <td>0.478</td>\n",
       "      <td>0.974</td>\n",
       "      <td>7.0</td>\n",
       "      <td>-3.866</td>\n",
       "      <td>1</td>\n",
       "      <td>0.1020</td>\n",
       "      <td>0.000266</td>\n",
       "      <td>0.098400</td>\n",
       "      <td>0.0738</td>\n",
       "      <td>0.268</td>\n",
       "      <td>141.077</td>\n",
       "      <td>248307.000000</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5994</th>\n",
       "      <td>35.0</td>\n",
       "      <td>0.456</td>\n",
       "      <td>0.264</td>\n",
       "      <td>6.0</td>\n",
       "      <td>-13.787</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0318</td>\n",
       "      <td>0.983000</td>\n",
       "      <td>0.343000</td>\n",
       "      <td>0.1180</td>\n",
       "      <td>0.694</td>\n",
       "      <td>94.352</td>\n",
       "      <td>4.722667</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12236 rows × 14 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Popularity  danceability  energy   key  loudness  mode  speechiness  \\\n",
       "9188         49.0         0.184   0.207   2.0   -17.119     1       0.0330   \n",
       "7086         69.0         0.553   0.872   1.0    -2.832     0       0.0782   \n",
       "3463         22.0         0.118   0.964   1.0    -4.060     1       0.4360   \n",
       "1897         58.0         0.634   0.648   3.0    -6.781     0       0.0412   \n",
       "1907         32.0         0.413   0.690  10.0    -6.868     1       0.0299   \n",
       "...           ...           ...     ...   ...       ...   ...          ...   \n",
       "9160         37.0         0.479   0.909   2.0    -7.404     1       0.0716   \n",
       "9859         71.0         0.671   0.256   7.0   -10.743     1       0.0387   \n",
       "11513        39.0         0.579   0.636   2.0    -8.574     0       0.0276   \n",
       "1688         66.0         0.478   0.974   7.0    -3.866     1       0.1020   \n",
       "5994         35.0         0.456   0.264   6.0   -13.787     1       0.0318   \n",
       "\n",
       "       acousticness  instrumentalness  liveness  valence    tempo  \\\n",
       "9188       0.940000          0.931000    0.0772    0.089   79.979   \n",
       "7086       0.198000          0.178129    0.0659    0.638  172.138   \n",
       "3463       0.000120          0.139000    0.3370    0.107  200.003   \n",
       "1897       0.361000          0.178129    0.6190    0.667   95.989   \n",
       "1907       0.212000          0.024100    0.3140    0.483   95.260   \n",
       "...             ...               ...       ...      ...      ...   \n",
       "9160       0.000091          0.758000    0.6340    0.278  111.892   \n",
       "9859       0.608000          0.178129    0.0737    0.694  115.871   \n",
       "11513      0.082100          0.922000    0.4460    0.898  132.652   \n",
       "1688       0.000266          0.098400    0.0738    0.268  141.077   \n",
       "5994       0.983000          0.343000    0.1180    0.694   94.352   \n",
       "\n",
       "       duration_in min/ms  time_signature  \n",
       "9188        251173.000000               3  \n",
       "7086             3.676217               4  \n",
       "3463        269787.000000               4  \n",
       "1897             4.130200               4  \n",
       "1907        267533.000000               4  \n",
       "...                   ...             ...  \n",
       "9160        310467.000000               4  \n",
       "9859        233560.000000               4  \n",
       "11513       280853.000000               4  \n",
       "1688        248307.000000               4  \n",
       "5994             4.722667               3  \n",
       "\n",
       "[12236 rows x 14 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create a python list of feature names\n",
    "feature_cols = ['Popularity', 'danceability', 'energy', 'key', 'loudness', 'mode','speechiness', 'acousticness',\n",
    "                'instrumentalness','liveness', 'valence', 'tempo', 'duration_in min/ms', 'time_signature']\n",
    "\n",
    "# Use the above feature name list to select the features from the DataFrame\n",
    "X = train_df[feature_cols]\n",
    "\n",
    "# Select a series of labels (the last column) from the DataFrame\n",
    "y = train_df['Class']\n",
    "\n",
    "# Randomly splitting the original dataset into training set and testing set\n",
    "# The function\"train_test_split\" from \"sklearn.cross_validation\" library performs random splitting.\n",
    "# \"test_size=0.15\" means that we pick 15% of data samples for testing set, and the rest (85%) for training set.\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.15, random_state = 3)\n",
    "\n",
    "print(X_train.shape)\n",
    "X_train"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "70094959",
   "metadata": {},
   "source": [
    "## Random Forest Implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "e5a9340c",
   "metadata": {},
   "outputs": [],
   "source": [
    "my_RandomForest = RandomForestClassifier(n_estimators=19, bootstrap=True, random_state=6)\n",
    "\n",
    "# Training ONLY on the training set:\n",
    "my_RandomForest.fit(X_train, y_train)\n",
    "\n",
    "# Testing on the testing set:\n",
    "y_predict = my_RandomForest.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "fa21b305",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The accuracy score of the Random Forest is: 0.7060185185185185\n"
     ]
    }
   ],
   "source": [
    "# We can now compare the \"predicted labels\" for the Testing Set with its \"actual labels\" to evaluate the accuracy \n",
    "# Function \"accuracy_score\" from \"sklearn.metrics\" will perform the element-to-element comparison and returns the \n",
    "# portion of correct predictions:\n",
    "\n",
    "score = accuracy_score(y_test, y_predict)\n",
    "\n",
    "print(\"The accuracy score of the Random Forest is: \" + str(score))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f9d9b146",
   "metadata": {},
   "source": [
    "### True Positive Rate (TPR) and False Positive Rate (FPR):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "7cc39932",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Estimating the probability (likelihood) of each label:\n",
    "y_predict_prob = my_RandomForest.predict_proba(X_test)\n",
    "\n",
    "fpr, tpr, thresholds = metrics.roc_curve(y_test, y_predict_prob[:,1], pos_label=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c8850668",
   "metadata": {},
   "source": [
    "### AUC:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "d26c5d25",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The AUC of the Random Forest is: 0.6946996742671012\n"
     ]
    }
   ],
   "source": [
    "AUC = metrics.auc(fpr, tpr)\n",
    "print(\"The AUC of the Random Forest is: \" + str(AUC))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc04c201",
   "metadata": {},
   "source": [
    "### ROC Curve:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "4338b3b5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "\n",
    "# Roc Curve:\n",
    "plt.plot(fpr, tpr, color='red', lw=2, label='ROC Curve (area = %0.2f)' % AUC)\n",
    "\n",
    "# Random Guess line:\n",
    "plt.plot([0, 1], [0, 1], color='blue', lw=1, linestyle='--')\n",
    "\n",
    "# Defining The Range of X-Axis and Y-Axis:\n",
    "plt.xlim([-0.005, 1.005])\n",
    "plt.ylim([0.0, 1.01])\n",
    "\n",
    "# Labels, Title, Legend:\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.legend(loc=\"lower right\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "75dd3adb",
   "metadata": {},
   "source": [
    "## Decision Tree Implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d91b5ad5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The accuracy score of the Decision Tree is: 0.649074074074074\n",
      "The AUC of the Decision Tree is: 0.5592052117263844\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "my_dt = DecisionTreeClassifier(random_state=3)\n",
    "\n",
    "#Training the model\n",
    "my_dt.fit(X_train, y_train)\n",
    "\n",
    "#Making predictions with test set\n",
    "y_predict_dt = my_dt.predict(X_test)\n",
    "\n",
    "#Calculating and displaying accuracy score of the decision tree\n",
    "score = accuracy_score(y_test, y_predict_dt)\n",
    "print(\"The accuracy score of the Decision Tree is: \" + str(score))\n",
    "\n",
    "#Calculating AUC\n",
    "y_predict_prob = my_dt.predict_proba(X_test)\n",
    "\n",
    "fpr, tpr, thresholds = metrics.roc_curve(y_test, y_predict_prob[:,1], pos_label=1)\n",
    "\n",
    "AUC = metrics.auc(fpr, tpr)\n",
    "print(\"The AUC of the Decision Tree is: \" + str(AUC))\n",
    "\n",
    "plt.figure()\n",
    "\n",
    "# Roc Curve:\n",
    "plt.plot(fpr, tpr, color='red', lw=2, \n",
    "         label='ROC Curve (area = %0.2f)' % AUC)\n",
    "\n",
    "# Random Guess line:\n",
    "plt.plot([0, 1], [0, 1], color='blue', lw=1, linestyle='--')\n",
    "\n",
    "# Defining The Range of X-Axis and Y-Axis:\n",
    "plt.xlim([-0.005, 1.005])\n",
    "plt.ylim([0.0, 1.01])\n",
    "\n",
    "# Labels, Title, Legend:\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.legend(loc=\"lower right\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ceacc717",
   "metadata": {},
   "source": [
    "## KNN Implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "14189bf6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The accuracy score of the knn is: 0.6944444444444444\n",
      "The AUC of the KNN is: 0.5829039739413682\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "k = 10\n",
    "my_knn = KNeighborsClassifier(n_neighbors=k)\n",
    "\n",
    "#Training the model\n",
    "my_knn.fit(X_train, y_train)\n",
    "\n",
    "#Making predictions with test set\n",
    "y_predict_knn = my_knn.predict(X_test)\n",
    "\n",
    "#Calculating and displaying accuracy score of the decision tree\n",
    "score = accuracy_score(y_test, y_predict_knn)\n",
    "print(\"The accuracy score of the knn is: \" + str(score))\n",
    "\n",
    "#Calculating AUC\n",
    "y_predict_prob = my_knn.predict_proba(X_test)\n",
    "\n",
    "fpr, tpr, thresholds = metrics.roc_curve(y_test, y_predict_prob[:,1], pos_label=1)\n",
    "\n",
    "AUC = metrics.auc(fpr, tpr)\n",
    "print(\"The AUC of the KNN is: \" + str(AUC))\n",
    "\n",
    "plt.figure()\n",
    "\n",
    "# Roc Curve:\n",
    "plt.plot(fpr, tpr, color='red', lw=2, \n",
    "         label='ROC Curve (area = %0.2f)' % AUC)\n",
    "\n",
    "# Random Guess line:\n",
    "plt.plot([0, 1], [0, 1], color='blue', lw=1, linestyle='--')\n",
    "\n",
    "# Defining The Range of X-Axis and Y-Axis:\n",
    "plt.xlim([-0.005, 1.005])\n",
    "plt.ylim([0.0, 1.01])\n",
    "\n",
    "# Labels, Title, Legend:\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.legend(loc=\"lower right\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c033d50c",
   "metadata": {},
   "source": [
    "# Testing accuracy and AUC if features with possible NaN values are just ignored"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "e63f9d40",
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_cols2 = ['danceability','energy','loudness','mode','speechiness','acousticness','liveness',\n",
    "                 'valence','tempo','duration_in min/ms', 'time_signature']\n",
    "\n",
    "X = train_df[feature_cols2]\n",
    "y = train_df['Class']\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.15, random_state = 3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1076a5e",
   "metadata": {},
   "source": [
    "## Random Forest Implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "bec0e7a8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The accuracy score of the Random Forest is: 0.6856481481481481\n",
      "The AUC of the Random Forest is: 0.6663379804560261\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "my_RandomForest = RandomForestClassifier(n_estimators=19, bootstrap=True, random_state=6)\n",
    "\n",
    "# Training ONLY on the training set:\n",
    "my_RandomForest.fit(X_train, y_train)\n",
    "\n",
    "# Testing on the testing set:\n",
    "y_predict = my_RandomForest.predict(X_test)\n",
    "\n",
    "#Calculating and displaying accuracy score of the random forest\n",
    "score = accuracy_score(y_test, y_predict)\n",
    "\n",
    "print(\"The accuracy score of the Random Forest is: \" + str(score))\n",
    "\n",
    "#Calculating AUC\n",
    "y_predict_prob = my_RandomForest.predict_proba(X_test)\n",
    "\n",
    "fpr, tpr, thresholds = metrics.roc_curve(y_test, y_predict_prob[:,1], pos_label=1)\n",
    "\n",
    "AUC = metrics.auc(fpr, tpr)\n",
    "print(\"The AUC of the Random Forest is: \" + str(AUC))\n",
    "\n",
    "plt.figure()\n",
    "\n",
    "# Roc Curve:\n",
    "plt.plot(fpr, tpr, color='red', lw=2, \n",
    "         label='ROC Curve (area = %0.2f)' % AUC)\n",
    "\n",
    "# Random Guess line:\n",
    "plt.plot([0, 1], [0, 1], color='blue', lw=1, linestyle='--')\n",
    "\n",
    "# Defining The Range of X-Axis and Y-Axis:\n",
    "plt.xlim([-0.005, 1.005])\n",
    "plt.ylim([0.0, 1.01])\n",
    "\n",
    "# Labels, Title, Legend:\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.legend(loc=\"lower right\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "84ffce08",
   "metadata": {},
   "source": [
    "## Decision Tree Implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "a32c4452",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The accuracy score of the Decision Tree is: 0.6638888888888889\n",
      "The AUC of the Decision Tree is: 0.5739309446254072\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "my_dt = DecisionTreeClassifier(random_state=3)\n",
    "\n",
    "#Training the model\n",
    "my_dt.fit(X_train, y_train)\n",
    "\n",
    "#Making predictions with test set\n",
    "y_predict_dt = my_dt.predict(X_test)\n",
    "\n",
    "#Calculating and displaying accuracy score of the decision tree\n",
    "score = accuracy_score(y_test, y_predict_dt)\n",
    "print(\"The accuracy score of the Decision Tree is: \" + str(score))\n",
    "\n",
    "#Calculating AUC\n",
    "y_predict_prob = my_dt.predict_proba(X_test)\n",
    "\n",
    "fpr, tpr, thresholds = metrics.roc_curve(y_test, y_predict_prob[:,1], pos_label=1)\n",
    "\n",
    "AUC = metrics.auc(fpr, tpr)\n",
    "print(\"The AUC of the Decision Tree is: \" + str(AUC))\n",
    "\n",
    "plt.figure()\n",
    "\n",
    "# Roc Curve:\n",
    "plt.plot(fpr, tpr, color='red', lw=2, \n",
    "         label='ROC Curve (area = %0.2f)' % AUC)\n",
    "\n",
    "# Random Guess line:\n",
    "plt.plot([0, 1], [0, 1], color='blue', lw=1, linestyle='--')\n",
    "\n",
    "# Defining The Range of X-Axis and Y-Axis:\n",
    "plt.xlim([-0.005, 1.005])\n",
    "plt.ylim([0.0, 1.01])\n",
    "\n",
    "# Labels, Title, Legend:\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.legend(loc=\"lower right\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b8be7a65",
   "metadata": {},
   "source": [
    "## KNN Implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "b6d09f05",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The accuracy score of KNN is: 0.6916666666666667\n",
      "The AUC of the KNN is: 0.575383973941368\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "k = 10\n",
    "my_knn = KNeighborsClassifier(n_neighbors=k)\n",
    "\n",
    "#Training the model\n",
    "my_knn.fit(X_train, y_train)\n",
    "\n",
    "#Making predictions with test set\n",
    "y_predict_knn = my_knn.predict(X_test)\n",
    "\n",
    "#Calculating and displaying accuracy score of knn classifier\n",
    "score = accuracy_score(y_test, y_predict_knn)\n",
    "print(\"The accuracy score of KNN is: \" + str(score))\n",
    "\n",
    "#Calculating AUC\n",
    "y_predict_prob = my_knn.predict_proba(X_test)\n",
    "\n",
    "fpr, tpr, thresholds = metrics.roc_curve(y_test, y_predict_prob[:,1], pos_label=1)\n",
    "\n",
    "AUC = metrics.auc(fpr, tpr)\n",
    "print(\"The AUC of the KNN is: \" + str(AUC))\n",
    "\n",
    "plt.figure()\n",
    "\n",
    "# Roc Curve:\n",
    "plt.plot(fpr, tpr, color='red', lw=2, \n",
    "         label='ROC Curve (area = %0.2f)' % AUC)\n",
    "\n",
    "# Random Guess line:\n",
    "plt.plot([0, 1], [0, 1], color='blue', lw=1, linestyle='--')\n",
    "\n",
    "# Defining The Range of X-Axis and Y-Axis:\n",
    "plt.xlim([-0.005, 1.005])\n",
    "plt.ylim([0.0, 1.01])\n",
    "\n",
    "# Labels, Title, Legend:\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.legend(loc=\"lower right\")\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}