19191_Marut_DSML_Project_Code.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Marut Priyadarshi\n",
    "#### DSE 2019\n",
    "#### 19191\n",
    "#### DSML Project"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.feature_selection import SelectKBest\n",
    "from sklearn.feature_selection import chi2\n",
    "from sklearn.ensemble import ExtraTreesClassifier\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.feature_selection import mutual_info_classif\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import train_test_split \n",
    "from sklearn import svm\n",
    "from sklearn.ensemble import RandomForestClassifier,BaggingClassifier\n",
    "from sklearn.naive_bayes import MultinomialNB, GaussianNB\n",
    "import pickle\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.preprocessing import StandardScaler,RobustScaler, MinMaxScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "x=pd.read_csv('music30s_trainin.csv')\n",
    "labels=pd.read_csv('music30s_trainlabel.csv')\n",
    "le=LabelEncoder()\n",
    "y=le.fit_transform(labels.iloc[:,1])\n",
    "yc=y\n",
    "y=pd.DataFrame({'label':y})\n",
    "data=pd.concat([x,y])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ASUS\\AppData\\Local\\Temp/ipykernel_8116/1200867342.py:2: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n",
      "  model.fit(x,y)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = ExtraTreesClassifier(random_state=42)\n",
    "model.fit(x,y)\n",
    "#use inbuilt class feature_importances of tree based classifiers\n",
    "#plot graph of feature importances for better visualization\n",
    "feat_importances = pd.Series(model.feature_importances_, index=x.columns)\n",
    "plt.figure(figsize=(10,10))\n",
    "feat_importances.nlargest(30).plot(kind='barh')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python39\\lib\\site-packages\\sklearn\\utils\\validation.py:63: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  return f(*args, **kwargs)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x1=pd.Series(mutual_info_classif(x,y, random_state=42), index=x.columns)\n",
    "plt.figure(figsize=(10,10))\n",
    "x1.nlargest(30).plot(kind='barh')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1440x1440 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "corrmat = data.corr()\n",
    "plt.figure(figsize=(20,20))\n",
    "sns.heatmap(corrmat,cmap=\"RdYlGn\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mfcc11_mean    1.000000\n",
      "mfcc13_mean    0.759363\n",
      "Name: mfcc11_mean, dtype: float64\n",
      "mfcc3_mean    1.0\n",
      "Name: mfcc3_mean, dtype: float64\n",
      "mfcc7_var    1.0\n",
      "Name: mfcc7_var, dtype: float64\n",
      "mfcc4_var    1.0\n",
      "Name: mfcc4_var, dtype: float64\n",
      "mfcc5_var    1.0\n",
      "Name: mfcc5_var, dtype: float64\n",
      "spectral_centroid_var     0.846002\n",
      "zero_crossing_rate_var    1.000000\n",
      "Name: zero_crossing_rate_var, dtype: float64\n",
      "mfcc1_var    1.0\n",
      "Name: mfcc1_var, dtype: float64\n",
      "mfcc17_mean    1.0\n",
      "Name: mfcc17_mean, dtype: float64\n",
      "mfcc5_mean    1.000000\n",
      "mfcc7_mean    0.763424\n",
      "Name: mfcc5_mean, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "c=x.corr()\n",
    "#looking for positive correlation\n",
    "#the features only occuring in feature importance\n",
    "print(c.loc['mfcc11_mean'][c.loc['mfcc11_mean'] > 0.75])\n",
    "print(c.loc['mfcc3_mean'][c.loc['mfcc3_mean'] > 0.75])\n",
    "print(c.loc['mfcc7_var'][c.loc['mfcc7_var'] > 0.75])\n",
    "print(c.loc['mfcc4_var'][c.loc['mfcc4_var'] > 0.75])\n",
    "print(c.loc['mfcc5_var'][c.loc['mfcc5_var'] > 0.75])\n",
    "# the features only occuring in mutual_info_classif\n",
    "print(c.loc['zero_crossing_rate_var'][c.loc['zero_crossing_rate_var'] > 0.75])\n",
    "print(c.loc['mfcc1_var'][c.loc['mfcc1_var'] > 0.75])\n",
    "print(c.loc['mfcc17_mean'][c.loc['mfcc17_mean'] > 0.75])\n",
    "print(c.loc['mfcc5_mean'][c.loc['mfcc5_mean'] > 0.75])\n",
    "\n",
    "#thus, we end up with 32 features out of 59 to finally train our classifier with."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Series([], Name: mfcc11_mean, dtype: float64)\n",
      "Series([], Name: mfcc3_mean, dtype: float64)\n",
      "Series([], Name: mfcc7_var, dtype: float64)\n",
      "Series([], Name: mfcc4_var, dtype: float64)\n",
      "Series([], Name: mfcc5_var, dtype: float64)\n",
      "Series([], Name: zero_crossing_rate_var, dtype: float64)\n",
      "Series([], Name: mfcc1_var, dtype: float64)\n",
      "Series([], Name: mfcc17_mean, dtype: float64)\n",
      "Series([], Name: mfcc5_mean, dtype: float64)\n"
     ]
    }
   ],
   "source": [
    "#checking for negative correlation\n",
    "\n",
    "#the features only occuring in feature importance\n",
    "print(c.loc['mfcc11_mean'][c.loc['mfcc11_mean'] < -0.75])\n",
    "print(c.loc['mfcc3_mean'][c.loc['mfcc3_mean'] < -0.75])\n",
    "print(c.loc['mfcc7_var'][c.loc['mfcc7_var'] < -0.75])\n",
    "print(c.loc['mfcc4_var'][c.loc['mfcc4_var'] < -0.75])\n",
    "print(c.loc['mfcc5_var'][c.loc['mfcc5_var'] < -0.75])\n",
    "# the features only occuring in mutual_info_classif\n",
    "print(c.loc['zero_crossing_rate_var'][c.loc['zero_crossing_rate_var'] < -0.75])\n",
    "print(c.loc['mfcc1_var'][c.loc['mfcc1_var'] < -0.75])\n",
    "print(c.loc['mfcc17_mean'][c.loc['mfcc17_mean'] < -0.75])\n",
    "print(c.loc['mfcc5_mean'][c.loc['mfcc5_mean'] < -0.75])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "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>chroma_stft_mean</th>\n",
       "      <th>length</th>\n",
       "      <th>perceptr_var</th>\n",
       "      <th>rolloff_mean</th>\n",
       "      <th>spectral_centroid_mean</th>\n",
       "      <th>spectral_bandwidth_mean</th>\n",
       "      <th>spectral_centroid_var</th>\n",
       "      <th>mfcc1_mean</th>\n",
       "      <th>rms_mean</th>\n",
       "      <th>rms_var</th>\n",
       "      <th>...</th>\n",
       "      <th>mfcc4_var</th>\n",
       "      <th>mfcc9_mean</th>\n",
       "      <th>harmony_var</th>\n",
       "      <th>mfcc11_mean</th>\n",
       "      <th>mfcc3_mean</th>\n",
       "      <th>mfcc7_var</th>\n",
       "      <th>mfcc4_var</th>\n",
       "      <th>mfcc5_var</th>\n",
       "      <th>mfcc1_var</th>\n",
       "      <th>mfcc17_mean</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.432846</td>\n",
       "      <td>661504</td>\n",
       "      <td>0.009514</td>\n",
       "      <td>6574.771380</td>\n",
       "      <td>3089.637185</td>\n",
       "      <td>2973.703815</td>\n",
       "      <td>1.311379e+06</td>\n",
       "      <td>-94.209412</td>\n",
       "      <td>0.142929</td>\n",
       "      <td>0.004758</td>\n",
       "      <td>...</td>\n",
       "      <td>506.841248</td>\n",
       "      <td>-6.113925</td>\n",
       "      <td>0.008987</td>\n",
       "      <td>0.332029</td>\n",
       "      <td>4.029281</td>\n",
       "      <td>189.074081</td>\n",
       "      <td>506.841248</td>\n",
       "      <td>393.235138</td>\n",
       "      <td>6272.106934</td>\n",
       "      <td>-5.005428</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.398761</td>\n",
       "      <td>661504</td>\n",
       "      <td>0.002661</td>\n",
       "      <td>4306.240937</td>\n",
       "      <td>2170.544505</td>\n",
       "      <td>2054.465986</td>\n",
       "      <td>2.470449e+05</td>\n",
       "      <td>-138.122055</td>\n",
       "      <td>0.092221</td>\n",
       "      <td>0.001716</td>\n",
       "      <td>...</td>\n",
       "      <td>261.225220</td>\n",
       "      <td>-2.105261</td>\n",
       "      <td>0.004687</td>\n",
       "      <td>4.206286</td>\n",
       "      <td>-35.343781</td>\n",
       "      <td>141.449295</td>\n",
       "      <td>261.225220</td>\n",
       "      <td>355.050415</td>\n",
       "      <td>2624.332764</td>\n",
       "      <td>-7.947048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.375874</td>\n",
       "      <td>661794</td>\n",
       "      <td>0.003022</td>\n",
       "      <td>4580.418732</td>\n",
       "      <td>2321.871533</td>\n",
       "      <td>2545.325000</td>\n",
       "      <td>5.433791e+05</td>\n",
       "      <td>-120.392059</td>\n",
       "      <td>0.099977</td>\n",
       "      <td>0.001147</td>\n",
       "      <td>...</td>\n",
       "      <td>298.257599</td>\n",
       "      <td>-14.665573</td>\n",
       "      <td>0.005084</td>\n",
       "      <td>-2.363255</td>\n",
       "      <td>-6.029607</td>\n",
       "      <td>80.607040</td>\n",
       "      <td>298.257599</td>\n",
       "      <td>226.056030</td>\n",
       "      <td>1672.984131</td>\n",
       "      <td>-3.584000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.424385</td>\n",
       "      <td>661794</td>\n",
       "      <td>0.010393</td>\n",
       "      <td>3828.288761</td>\n",
       "      <td>1886.954809</td>\n",
       "      <td>1913.258610</td>\n",
       "      <td>2.533910e+05</td>\n",
       "      <td>-92.703461</td>\n",
       "      <td>0.175975</td>\n",
       "      <td>0.003569</td>\n",
       "      <td>...</td>\n",
       "      <td>331.039490</td>\n",
       "      <td>-8.203340</td>\n",
       "      <td>0.014277</td>\n",
       "      <td>-5.647607</td>\n",
       "      <td>-35.067982</td>\n",
       "      <td>149.097458</td>\n",
       "      <td>331.039490</td>\n",
       "      <td>301.549194</td>\n",
       "      <td>3131.187744</td>\n",
       "      <td>-6.554968</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.456134</td>\n",
       "      <td>661794</td>\n",
       "      <td>0.002530</td>\n",
       "      <td>4186.459372</td>\n",
       "      <td>2269.128009</td>\n",
       "      <td>1928.953538</td>\n",
       "      <td>1.357942e+05</td>\n",
       "      <td>-100.384117</td>\n",
       "      <td>0.097609</td>\n",
       "      <td>0.000364</td>\n",
       "      <td>...</td>\n",
       "      <td>113.169205</td>\n",
       "      <td>-19.065331</td>\n",
       "      <td>0.004224</td>\n",
       "      <td>-13.772540</td>\n",
       "      <td>-57.247860</td>\n",
       "      <td>110.296234</td>\n",
       "      <td>113.169205</td>\n",
       "      <td>93.510063</td>\n",
       "      <td>1325.951050</td>\n",
       "      <td>-12.668804</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",
       "      <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>895</th>\n",
       "      <td>0.469879</td>\n",
       "      <td>661504</td>\n",
       "      <td>0.009818</td>\n",
       "      <td>5796.886499</td>\n",
       "      <td>2868.233682</td>\n",
       "      <td>2676.125867</td>\n",
       "      <td>1.353624e+06</td>\n",
       "      <td>-127.277176</td>\n",
       "      <td>0.144261</td>\n",
       "      <td>0.008266</td>\n",
       "      <td>...</td>\n",
       "      <td>266.928986</td>\n",
       "      <td>4.607176</td>\n",
       "      <td>0.013576</td>\n",
       "      <td>5.581950</td>\n",
       "      <td>4.438226</td>\n",
       "      <td>146.197754</td>\n",
       "      <td>266.928986</td>\n",
       "      <td>186.883072</td>\n",
       "      <td>4943.034180</td>\n",
       "      <td>-5.572653</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>896</th>\n",
       "      <td>0.290960</td>\n",
       "      <td>661794</td>\n",
       "      <td>0.001711</td>\n",
       "      <td>2243.341758</td>\n",
       "      <td>1109.352784</td>\n",
       "      <td>1463.217702</td>\n",
       "      <td>1.386997e+05</td>\n",
       "      <td>-249.284561</td>\n",
       "      <td>0.089215</td>\n",
       "      <td>0.003113</td>\n",
       "      <td>...</td>\n",
       "      <td>298.369965</td>\n",
       "      <td>1.412567</td>\n",
       "      <td>0.006751</td>\n",
       "      <td>-0.981921</td>\n",
       "      <td>0.671435</td>\n",
       "      <td>96.463455</td>\n",
       "      <td>298.369965</td>\n",
       "      <td>183.418121</td>\n",
       "      <td>3980.260986</td>\n",
       "      <td>-2.823469</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>897</th>\n",
       "      <td>0.454745</td>\n",
       "      <td>661504</td>\n",
       "      <td>0.025124</td>\n",
       "      <td>6616.305648</td>\n",
       "      <td>3123.151410</td>\n",
       "      <td>2970.246872</td>\n",
       "      <td>9.471130e+05</td>\n",
       "      <td>-42.758705</td>\n",
       "      <td>0.223571</td>\n",
       "      <td>0.013054</td>\n",
       "      <td>...</td>\n",
       "      <td>345.183136</td>\n",
       "      <td>3.192147</td>\n",
       "      <td>0.018758</td>\n",
       "      <td>-4.768050</td>\n",
       "      <td>4.770786</td>\n",
       "      <td>198.291672</td>\n",
       "      <td>345.183136</td>\n",
       "      <td>275.695343</td>\n",
       "      <td>4607.766113</td>\n",
       "      <td>-1.865332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>898</th>\n",
       "      <td>0.437934</td>\n",
       "      <td>661794</td>\n",
       "      <td>0.016427</td>\n",
       "      <td>4003.377184</td>\n",
       "      <td>1821.519761</td>\n",
       "      <td>2522.060569</td>\n",
       "      <td>7.452975e+05</td>\n",
       "      <td>-118.051056</td>\n",
       "      <td>0.252890</td>\n",
       "      <td>0.013661</td>\n",
       "      <td>...</td>\n",
       "      <td>281.339478</td>\n",
       "      <td>6.773134</td>\n",
       "      <td>0.042239</td>\n",
       "      <td>8.584536</td>\n",
       "      <td>20.755558</td>\n",
       "      <td>107.842049</td>\n",
       "      <td>281.339478</td>\n",
       "      <td>252.895935</td>\n",
       "      <td>4448.062012</td>\n",
       "      <td>-5.466094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>899</th>\n",
       "      <td>0.487560</td>\n",
       "      <td>661794</td>\n",
       "      <td>0.010166</td>\n",
       "      <td>5228.721360</td>\n",
       "      <td>2718.609163</td>\n",
       "      <td>2271.378022</td>\n",
       "      <td>2.446342e+05</td>\n",
       "      <td>-24.949142</td>\n",
       "      <td>0.207197</td>\n",
       "      <td>0.001459</td>\n",
       "      <td>...</td>\n",
       "      <td>168.014450</td>\n",
       "      <td>-11.454040</td>\n",
       "      <td>0.021699</td>\n",
       "      <td>-14.156014</td>\n",
       "      <td>-30.369211</td>\n",
       "      <td>108.762215</td>\n",
       "      <td>168.014450</td>\n",
       "      <td>88.567520</td>\n",
       "      <td>1509.886353</td>\n",
       "      <td>-6.626670</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>900 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     chroma_stft_mean  length  perceptr_var  rolloff_mean  \\\n",
       "0            0.432846  661504      0.009514   6574.771380   \n",
       "1            0.398761  661504      0.002661   4306.240937   \n",
       "2            0.375874  661794      0.003022   4580.418732   \n",
       "3            0.424385  661794      0.010393   3828.288761   \n",
       "4            0.456134  661794      0.002530   4186.459372   \n",
       "..                ...     ...           ...           ...   \n",
       "895          0.469879  661504      0.009818   5796.886499   \n",
       "896          0.290960  661794      0.001711   2243.341758   \n",
       "897          0.454745  661504      0.025124   6616.305648   \n",
       "898          0.437934  661794      0.016427   4003.377184   \n",
       "899          0.487560  661794      0.010166   5228.721360   \n",
       "\n",
       "     spectral_centroid_mean  spectral_bandwidth_mean  spectral_centroid_var  \\\n",
       "0               3089.637185              2973.703815           1.311379e+06   \n",
       "1               2170.544505              2054.465986           2.470449e+05   \n",
       "2               2321.871533              2545.325000           5.433791e+05   \n",
       "3               1886.954809              1913.258610           2.533910e+05   \n",
       "4               2269.128009              1928.953538           1.357942e+05   \n",
       "..                      ...                      ...                    ...   \n",
       "895             2868.233682              2676.125867           1.353624e+06   \n",
       "896             1109.352784              1463.217702           1.386997e+05   \n",
       "897             3123.151410              2970.246872           9.471130e+05   \n",
       "898             1821.519761              2522.060569           7.452975e+05   \n",
       "899             2718.609163              2271.378022           2.446342e+05   \n",
       "\n",
       "     mfcc1_mean  rms_mean   rms_var  ...   mfcc4_var  mfcc9_mean  harmony_var  \\\n",
       "0    -94.209412  0.142929  0.004758  ...  506.841248   -6.113925     0.008987   \n",
       "1   -138.122055  0.092221  0.001716  ...  261.225220   -2.105261     0.004687   \n",
       "2   -120.392059  0.099977  0.001147  ...  298.257599  -14.665573     0.005084   \n",
       "3    -92.703461  0.175975  0.003569  ...  331.039490   -8.203340     0.014277   \n",
       "4   -100.384117  0.097609  0.000364  ...  113.169205  -19.065331     0.004224   \n",
       "..          ...       ...       ...  ...         ...         ...          ...   \n",
       "895 -127.277176  0.144261  0.008266  ...  266.928986    4.607176     0.013576   \n",
       "896 -249.284561  0.089215  0.003113  ...  298.369965    1.412567     0.006751   \n",
       "897  -42.758705  0.223571  0.013054  ...  345.183136    3.192147     0.018758   \n",
       "898 -118.051056  0.252890  0.013661  ...  281.339478    6.773134     0.042239   \n",
       "899  -24.949142  0.207197  0.001459  ...  168.014450  -11.454040     0.021699   \n",
       "\n",
       "     mfcc11_mean  mfcc3_mean   mfcc7_var   mfcc4_var   mfcc5_var    mfcc1_var  \\\n",
       "0       0.332029    4.029281  189.074081  506.841248  393.235138  6272.106934   \n",
       "1       4.206286  -35.343781  141.449295  261.225220  355.050415  2624.332764   \n",
       "2      -2.363255   -6.029607   80.607040  298.257599  226.056030  1672.984131   \n",
       "3      -5.647607  -35.067982  149.097458  331.039490  301.549194  3131.187744   \n",
       "4     -13.772540  -57.247860  110.296234  113.169205   93.510063  1325.951050   \n",
       "..           ...         ...         ...         ...         ...          ...   \n",
       "895     5.581950    4.438226  146.197754  266.928986  186.883072  4943.034180   \n",
       "896    -0.981921    0.671435   96.463455  298.369965  183.418121  3980.260986   \n",
       "897    -4.768050    4.770786  198.291672  345.183136  275.695343  4607.766113   \n",
       "898     8.584536   20.755558  107.842049  281.339478  252.895935  4448.062012   \n",
       "899   -14.156014  -30.369211  108.762215  168.014450   88.567520  1509.886353   \n",
       "\n",
       "     mfcc17_mean  \n",
       "0      -5.005428  \n",
       "1      -7.947048  \n",
       "2      -3.584000  \n",
       "3      -6.554968  \n",
       "4     -12.668804  \n",
       "..           ...  \n",
       "895    -5.572653  \n",
       "896    -2.823469  \n",
       "897    -1.865332  \n",
       "898    -5.466094  \n",
       "899    -6.626670  \n",
       "\n",
       "[900 rows x 32 columns]"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# making a dataframe of all the selected features\n",
    "xc=x[['chroma_stft_mean','length','perceptr_var','rolloff_mean','spectral_centroid_mean','spectral_bandwidth_mean','spectral_centroid_var','mfcc1_mean','rms_mean','rms_var','mfcc2_mean','mfcc4_mean','rolloff_var','mfcc12_mean','mfcc6_var','chroma_stft_var','mfcc20_var','mfcc8_mean','mfcc7_mean','perceptr_mean','mfcc6_mean','zero_crossing_rate_mean','mfcc4_var','mfcc9_mean','harmony_var','mfcc11_mean','mfcc3_mean','mfcc7_var','mfcc4_var','mfcc5_var','mfcc1_var','mfcc17_mean']].copy()\n",
    "xc"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Standard Scaled Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "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>chroma_stft_mean</th>\n",
       "      <th>length</th>\n",
       "      <th>perceptr_var</th>\n",
       "      <th>rolloff_mean</th>\n",
       "      <th>spectral_centroid_mean</th>\n",
       "      <th>spectral_bandwidth_mean</th>\n",
       "      <th>spectral_centroid_var</th>\n",
       "      <th>mfcc1_mean</th>\n",
       "      <th>rms_mean</th>\n",
       "      <th>rms_var</th>\n",
       "      <th>...</th>\n",
       "      <th>mfcc4_var</th>\n",
       "      <th>mfcc9_mean</th>\n",
       "      <th>harmony_var</th>\n",
       "      <th>mfcc11_mean</th>\n",
       "      <th>mfcc3_mean</th>\n",
       "      <th>mfcc7_var</th>\n",
       "      <th>mfcc4_var</th>\n",
       "      <th>mfcc5_var</th>\n",
       "      <th>mfcc1_var</th>\n",
       "      <th>mfcc17_mean</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.642278</td>\n",
       "      <td>-0.298576</td>\n",
       "      <td>0.603005</td>\n",
       "      <td>1.266091</td>\n",
       "      <td>1.233038</td>\n",
       "      <td>1.386263</td>\n",
       "      <td>2.110851</td>\n",
       "      <td>0.493279</td>\n",
       "      <td>0.177059</td>\n",
       "      <td>0.481112</td>\n",
       "      <td>...</td>\n",
       "      <td>2.529018</td>\n",
       "      <td>0.101920</td>\n",
       "      <td>-0.306906</td>\n",
       "      <td>0.943894</td>\n",
       "      <td>0.596189</td>\n",
       "      <td>1.292123</td>\n",
       "      <td>2.529018</td>\n",
       "      <td>2.161359</td>\n",
       "      <td>0.907146</td>\n",
       "      <td>-0.222530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.227633</td>\n",
       "      <td>-0.298576</td>\n",
       "      <td>-0.463828</td>\n",
       "      <td>-0.174821</td>\n",
       "      <td>-0.050038</td>\n",
       "      <td>-0.362538</td>\n",
       "      <td>-0.550743</td>\n",
       "      <td>0.052394</td>\n",
       "      <td>-0.595372</td>\n",
       "      <td>-0.362168</td>\n",
       "      <td>...</td>\n",
       "      <td>0.369853</td>\n",
       "      <td>0.585031</td>\n",
       "      <td>-0.673346</td>\n",
       "      <td>1.519104</td>\n",
       "      <td>-1.222094</td>\n",
       "      <td>0.456196</td>\n",
       "      <td>0.369853</td>\n",
       "      <td>1.784297</td>\n",
       "      <td>-0.392146</td>\n",
       "      <td>-0.877581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.050788</td>\n",
       "      <td>-0.142344</td>\n",
       "      <td>-0.407656</td>\n",
       "      <td>-0.000670</td>\n",
       "      <td>0.161218</td>\n",
       "      <td>0.571295</td>\n",
       "      <td>0.190304</td>\n",
       "      <td>0.230404</td>\n",
       "      <td>-0.477223</td>\n",
       "      <td>-0.519769</td>\n",
       "      <td>...</td>\n",
       "      <td>0.695398</td>\n",
       "      <td>-0.928696</td>\n",
       "      <td>-0.639552</td>\n",
       "      <td>0.543725</td>\n",
       "      <td>0.131661</td>\n",
       "      <td>-0.611727</td>\n",
       "      <td>0.695398</td>\n",
       "      <td>0.510521</td>\n",
       "      <td>-0.731004</td>\n",
       "      <td>0.093998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.539357</td>\n",
       "      <td>-0.142344</td>\n",
       "      <td>0.739850</td>\n",
       "      <td>-0.478404</td>\n",
       "      <td>-0.445936</td>\n",
       "      <td>-0.631178</td>\n",
       "      <td>-0.534873</td>\n",
       "      <td>0.508399</td>\n",
       "      <td>0.680464</td>\n",
       "      <td>0.151445</td>\n",
       "      <td>...</td>\n",
       "      <td>0.983578</td>\n",
       "      <td>-0.149889</td>\n",
       "      <td>0.143906</td>\n",
       "      <td>0.056098</td>\n",
       "      <td>-1.209357</td>\n",
       "      <td>0.590440</td>\n",
       "      <td>0.983578</td>\n",
       "      <td>1.255991</td>\n",
       "      <td>-0.211610</td>\n",
       "      <td>-0.567588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.925586</td>\n",
       "      <td>-0.142344</td>\n",
       "      <td>-0.484184</td>\n",
       "      <td>-0.250903</td>\n",
       "      <td>0.087587</td>\n",
       "      <td>-0.601319</td>\n",
       "      <td>-0.828949</td>\n",
       "      <td>0.431285</td>\n",
       "      <td>-0.513290</td>\n",
       "      <td>-0.736831</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.931680</td>\n",
       "      <td>-1.458940</td>\n",
       "      <td>-0.712798</td>\n",
       "      <td>-1.150210</td>\n",
       "      <td>-2.233644</td>\n",
       "      <td>-0.090613</td>\n",
       "      <td>-0.931680</td>\n",
       "      <td>-0.798326</td>\n",
       "      <td>-0.854613</td>\n",
       "      <td>-1.929039</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",
       "      <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>895</th>\n",
       "      <td>1.092797</td>\n",
       "      <td>-0.298576</td>\n",
       "      <td>0.650282</td>\n",
       "      <td>0.771998</td>\n",
       "      <td>0.923953</td>\n",
       "      <td>0.820137</td>\n",
       "      <td>2.216495</td>\n",
       "      <td>0.161277</td>\n",
       "      <td>0.197356</td>\n",
       "      <td>1.453363</td>\n",
       "      <td>...</td>\n",
       "      <td>0.419994</td>\n",
       "      <td>1.393991</td>\n",
       "      <td>0.084195</td>\n",
       "      <td>1.723349</td>\n",
       "      <td>0.615075</td>\n",
       "      <td>0.539543</td>\n",
       "      <td>0.419994</td>\n",
       "      <td>0.123701</td>\n",
       "      <td>0.433747</td>\n",
       "      <td>-0.348842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>896</th>\n",
       "      <td>-1.083787</td>\n",
       "      <td>-0.142344</td>\n",
       "      <td>-0.611628</td>\n",
       "      <td>-1.485121</td>\n",
       "      <td>-1.531488</td>\n",
       "      <td>-1.487357</td>\n",
       "      <td>-0.821683</td>\n",
       "      <td>-1.063685</td>\n",
       "      <td>-0.641160</td>\n",
       "      <td>0.025120</td>\n",
       "      <td>...</td>\n",
       "      <td>0.696386</td>\n",
       "      <td>1.008987</td>\n",
       "      <td>-0.497451</td>\n",
       "      <td>0.748812</td>\n",
       "      <td>0.441121</td>\n",
       "      <td>-0.333410</td>\n",
       "      <td>0.696386</td>\n",
       "      <td>0.089486</td>\n",
       "      <td>0.090819</td>\n",
       "      <td>0.263356</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>897</th>\n",
       "      <td>0.908685</td>\n",
       "      <td>-0.298576</td>\n",
       "      <td>3.032946</td>\n",
       "      <td>1.292472</td>\n",
       "      <td>1.279825</td>\n",
       "      <td>1.379687</td>\n",
       "      <td>1.199927</td>\n",
       "      <td>1.009848</td>\n",
       "      <td>1.405492</td>\n",
       "      <td>2.780412</td>\n",
       "      <td>...</td>\n",
       "      <td>1.107912</td>\n",
       "      <td>1.223456</td>\n",
       "      <td>0.525843</td>\n",
       "      <td>0.186686</td>\n",
       "      <td>0.630432</td>\n",
       "      <td>1.453913</td>\n",
       "      <td>1.107912</td>\n",
       "      <td>1.000692</td>\n",
       "      <td>0.314328</td>\n",
       "      <td>0.476718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>898</th>\n",
       "      <td>0.704175</td>\n",
       "      <td>-0.142344</td>\n",
       "      <td>1.679062</td>\n",
       "      <td>-0.367192</td>\n",
       "      <td>-0.537285</td>\n",
       "      <td>0.527036</td>\n",
       "      <td>0.695244</td>\n",
       "      <td>0.253908</td>\n",
       "      <td>1.852109</td>\n",
       "      <td>2.948690</td>\n",
       "      <td>...</td>\n",
       "      <td>0.546674</td>\n",
       "      <td>1.655025</td>\n",
       "      <td>2.526927</td>\n",
       "      <td>2.169143</td>\n",
       "      <td>1.368623</td>\n",
       "      <td>-0.133689</td>\n",
       "      <td>0.546674</td>\n",
       "      <td>0.775556</td>\n",
       "      <td>0.257444</td>\n",
       "      <td>-0.325113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>899</th>\n",
       "      <td>1.307879</td>\n",
       "      <td>-0.142344</td>\n",
       "      <td>0.704405</td>\n",
       "      <td>0.411115</td>\n",
       "      <td>0.715074</td>\n",
       "      <td>0.050125</td>\n",
       "      <td>-0.556771</td>\n",
       "      <td>1.188657</td>\n",
       "      <td>1.156067</td>\n",
       "      <td>-0.433309</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.449545</td>\n",
       "      <td>-0.541653</td>\n",
       "      <td>0.776437</td>\n",
       "      <td>-1.207145</td>\n",
       "      <td>-0.992364</td>\n",
       "      <td>-0.117538</td>\n",
       "      <td>-0.449545</td>\n",
       "      <td>-0.847132</td>\n",
       "      <td>-0.789097</td>\n",
       "      <td>-0.583554</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>900 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     chroma_stft_mean    length  perceptr_var  rolloff_mean  \\\n",
       "0            0.642278 -0.298576      0.603005      1.266091   \n",
       "1            0.227633 -0.298576     -0.463828     -0.174821   \n",
       "2           -0.050788 -0.142344     -0.407656     -0.000670   \n",
       "3            0.539357 -0.142344      0.739850     -0.478404   \n",
       "4            0.925586 -0.142344     -0.484184     -0.250903   \n",
       "..                ...       ...           ...           ...   \n",
       "895          1.092797 -0.298576      0.650282      0.771998   \n",
       "896         -1.083787 -0.142344     -0.611628     -1.485121   \n",
       "897          0.908685 -0.298576      3.032946      1.292472   \n",
       "898          0.704175 -0.142344      1.679062     -0.367192   \n",
       "899          1.307879 -0.142344      0.704405      0.411115   \n",
       "\n",
       "     spectral_centroid_mean  spectral_bandwidth_mean  spectral_centroid_var  \\\n",
       "0                  1.233038                 1.386263               2.110851   \n",
       "1                 -0.050038                -0.362538              -0.550743   \n",
       "2                  0.161218                 0.571295               0.190304   \n",
       "3                 -0.445936                -0.631178              -0.534873   \n",
       "4                  0.087587                -0.601319              -0.828949   \n",
       "..                      ...                      ...                    ...   \n",
       "895                0.923953                 0.820137               2.216495   \n",
       "896               -1.531488                -1.487357              -0.821683   \n",
       "897                1.279825                 1.379687               1.199927   \n",
       "898               -0.537285                 0.527036               0.695244   \n",
       "899                0.715074                 0.050125              -0.556771   \n",
       "\n",
       "     mfcc1_mean  rms_mean   rms_var  ...  mfcc4_var  mfcc9_mean  harmony_var  \\\n",
       "0      0.493279  0.177059  0.481112  ...   2.529018    0.101920    -0.306906   \n",
       "1      0.052394 -0.595372 -0.362168  ...   0.369853    0.585031    -0.673346   \n",
       "2      0.230404 -0.477223 -0.519769  ...   0.695398   -0.928696    -0.639552   \n",
       "3      0.508399  0.680464  0.151445  ...   0.983578   -0.149889     0.143906   \n",
       "4      0.431285 -0.513290 -0.736831  ...  -0.931680   -1.458940    -0.712798   \n",
       "..          ...       ...       ...  ...        ...         ...          ...   \n",
       "895    0.161277  0.197356  1.453363  ...   0.419994    1.393991     0.084195   \n",
       "896   -1.063685 -0.641160  0.025120  ...   0.696386    1.008987    -0.497451   \n",
       "897    1.009848  1.405492  2.780412  ...   1.107912    1.223456     0.525843   \n",
       "898    0.253908  1.852109  2.948690  ...   0.546674    1.655025     2.526927   \n",
       "899    1.188657  1.156067 -0.433309  ...  -0.449545   -0.541653     0.776437   \n",
       "\n",
       "     mfcc11_mean  mfcc3_mean  mfcc7_var  mfcc4_var  mfcc5_var  mfcc1_var  \\\n",
       "0       0.943894    0.596189   1.292123   2.529018   2.161359   0.907146   \n",
       "1       1.519104   -1.222094   0.456196   0.369853   1.784297  -0.392146   \n",
       "2       0.543725    0.131661  -0.611727   0.695398   0.510521  -0.731004   \n",
       "3       0.056098   -1.209357   0.590440   0.983578   1.255991  -0.211610   \n",
       "4      -1.150210   -2.233644  -0.090613  -0.931680  -0.798326  -0.854613   \n",
       "..           ...         ...        ...        ...        ...        ...   \n",
       "895     1.723349    0.615075   0.539543   0.419994   0.123701   0.433747   \n",
       "896     0.748812    0.441121  -0.333410   0.696386   0.089486   0.090819   \n",
       "897     0.186686    0.630432   1.453913   1.107912   1.000692   0.314328   \n",
       "898     2.169143    1.368623  -0.133689   0.546674   0.775556   0.257444   \n",
       "899    -1.207145   -0.992364  -0.117538  -0.449545  -0.847132  -0.789097   \n",
       "\n",
       "     mfcc17_mean  \n",
       "0      -0.222530  \n",
       "1      -0.877581  \n",
       "2       0.093998  \n",
       "3      -0.567588  \n",
       "4      -1.929039  \n",
       "..           ...  \n",
       "895    -0.348842  \n",
       "896     0.263356  \n",
       "897     0.476718  \n",
       "898    -0.325113  \n",
       "899    -0.583554  \n",
       "\n",
       "[900 rows x 32 columns]"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler = StandardScaler()\n",
    "scaler.fit(xc)\n",
    "xct=pd.DataFrame(scaler.transform(xc),columns=xc.columns)\n",
    "scaler.fit(x)\n",
    "xt=pd.DataFrame(scaler.transform(x),columns=x.columns)\n",
    "xct\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train, x_test, y_train, y_test = train_test_split(xct, yc, test_size=0.20, random_state=42, stratify=yc)\n",
    "xo_train, xo_test, yo_train, yo_test = train_test_split(xt, yc, test_size=0.20, random_state=42, stratify=yc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### K-Nearest Neighbour"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6796454045009795 {'leaf_size': 11, 'n_jobs': -1, 'n_neighbors': 5, 'p': 1}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.611111</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.400000</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>0.631579</td>\n",
       "      <td>0.562500</td>\n",
       "      <td>0.416667</td>\n",
       "      <td>0.638889</td>\n",
       "      <td>0.674567</td>\n",
       "      <td>0.673486</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.647059</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.388889</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.263158</td>\n",
       "      <td>0.638889</td>\n",
       "      <td>0.641022</td>\n",
       "      <td>0.638889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.628571</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.528302</td>\n",
       "      <td>0.697674</td>\n",
       "      <td>0.518519</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.648649</td>\n",
       "      <td>0.529412</td>\n",
       "      <td>0.322581</td>\n",
       "      <td>0.638889</td>\n",
       "      <td>0.638760</td>\n",
       "      <td>0.637060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.638889</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.611111   0.888889   0.400000   0.600000   0.777778   1.000000   \n",
       "recall      0.647059   0.888889   0.777778   0.833333   0.388889   0.777778   \n",
       "f1-score    0.628571   0.888889   0.528302   0.697674   0.518519   0.875000   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.857143   0.631579   0.562500   0.416667  0.638889    0.674567   \n",
       "recall      0.666667   0.666667   0.500000   0.263158  0.638889    0.641022   \n",
       "f1-score    0.750000   0.648649   0.529412   0.322581  0.638889    0.638760   \n",
       "support    18.000000  18.000000  18.000000  19.000000  0.638889  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.673486  \n",
       "recall         0.638889  \n",
       "f1-score       0.637060  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with extracted features\n",
    "parameters={'p':[1,2,3],'n_jobs':[-1],'n_neighbors':[3,4,5,6,7],'leaf_size':[11,12,13,14,15,16]}\n",
    "knn=KNeighborsClassifier()\n",
    "clf1 = GridSearchCV(knn, parameters, cv=5, scoring='f1_macro')\n",
    "clf1.fit(x_train,y_train)\n",
    "print(clf1.best_score_,clf1.best_params_)\n",
    "y_pred=clf1.predict(x_test)\n",
    "report=pd.DataFrame(classification_report(y_test,y_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6630146284039693 {'leaf_size': 11, 'n_jobs': -1, 'n_neighbors': 5, 'p': 1}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.714286</td>\n",
       "      <td>0.894737</td>\n",
       "      <td>0.464286</td>\n",
       "      <td>0.571429</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.928571</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>0.476190</td>\n",
       "      <td>0.315789</td>\n",
       "      <td>0.627778</td>\n",
       "      <td>0.648751</td>\n",
       "      <td>0.646537</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.588235</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.277778</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.315789</td>\n",
       "      <td>0.627778</td>\n",
       "      <td>0.629291</td>\n",
       "      <td>0.627778</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.645161</td>\n",
       "      <td>0.918919</td>\n",
       "      <td>0.565217</td>\n",
       "      <td>0.615385</td>\n",
       "      <td>0.370370</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.787879</td>\n",
       "      <td>0.736842</td>\n",
       "      <td>0.512821</td>\n",
       "      <td>0.315789</td>\n",
       "      <td>0.627778</td>\n",
       "      <td>0.628088</td>\n",
       "      <td>0.626259</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.627778</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.714286   0.894737   0.464286   0.571429   0.555556   0.928571   \n",
       "recall      0.588235   0.944444   0.722222   0.666667   0.277778   0.722222   \n",
       "f1-score    0.645161   0.918919   0.565217   0.615385   0.370370   0.812500   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.866667   0.700000   0.476190   0.315789  0.627778    0.648751   \n",
       "recall      0.722222   0.777778   0.555556   0.315789  0.627778    0.629291   \n",
       "f1-score    0.787879   0.736842   0.512821   0.315789  0.627778    0.628088   \n",
       "support    18.000000  18.000000  18.000000  19.000000  0.627778  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.646537  \n",
       "recall         0.627778  \n",
       "f1-score       0.626259  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with all features\n",
    "parameters={'p':[1,2,3],'n_jobs':[-1],'n_neighbors':[3,5,7,9],'leaf_size':[11,12,13,14,15,16]}\n",
    "knn=KNeighborsClassifier()\n",
    "clf1o = GridSearchCV(knn, parameters, cv=5, scoring='f1_macro')\n",
    "clf1o.fit(xo_train,yo_train)\n",
    "print(clf1o.best_score_,clf1o.best_params_)\n",
    "yo_pred=clf1o.predict(xo_test)\n",
    "report=pd.DataFrame(classification_report(yo_test,yo_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Support Vector Machine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.745487003810881 {'C': 7.97, 'kernel': 'rbf'}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.800000</td>\n",
       "      <td>0.937500</td>\n",
       "      <td>0.736842</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.937500</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>0.619048</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.541667</td>\n",
       "      <td>0.716667</td>\n",
       "      <td>0.732533</td>\n",
       "      <td>0.731098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.705882</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.684211</td>\n",
       "      <td>0.716667</td>\n",
       "      <td>0.716787</td>\n",
       "      <td>0.716667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.882353</td>\n",
       "      <td>0.756757</td>\n",
       "      <td>0.606061</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.882353</td>\n",
       "      <td>0.823529</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.631579</td>\n",
       "      <td>0.604651</td>\n",
       "      <td>0.716667</td>\n",
       "      <td>0.721506</td>\n",
       "      <td>0.720699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.716667</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.800000   0.937500   0.736842   0.666667   0.611111   0.937500   \n",
       "recall      0.705882   0.833333   0.777778   0.555556   0.611111   0.833333   \n",
       "f1-score    0.750000   0.882353   0.756757   0.606061   0.611111   0.882353   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.875000   0.619048   0.600000   0.541667  0.716667    0.732533   \n",
       "recall      0.777778   0.722222   0.666667   0.684211  0.716667    0.716787   \n",
       "f1-score    0.823529   0.666667   0.631579   0.604651  0.716667    0.721506   \n",
       "support    18.000000  18.000000  18.000000  19.000000  0.716667  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.731098  \n",
       "recall         0.716667  \n",
       "f1-score       0.720699  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with extracted features\n",
    "parameters = {'kernel':['rbf'], 'C':[7.96,7.97,7.98,7.99]}\n",
    "svc=svm.SVC()\n",
    "clf2 = GridSearchCV(svc, parameters, cv=5, scoring='f1_macro')\n",
    "clf2.fit(x_train,y_train)\n",
    "print(clf2.best_score_,clf2.best_params_)\n",
    "y_pred=clf2.predict(x_test)\n",
    "report=pd.DataFrame(classification_report(y_test,y_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7273241773874111 {'C': 7.89, 'kernel': 'rbf'}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.764706</td>\n",
       "      <td>0.894737</td>\n",
       "      <td>0.789474</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.714286</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.590909</td>\n",
       "      <td>0.571429</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.756860</td>\n",
       "      <td>0.755786</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.764706</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.631579</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.750740</td>\n",
       "      <td>0.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.764706</td>\n",
       "      <td>0.918919</td>\n",
       "      <td>0.810811</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.625000</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.823529</td>\n",
       "      <td>0.764706</td>\n",
       "      <td>0.650000</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.751323</td>\n",
       "      <td>0.750408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.75</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.764706   0.894737   0.789474   0.722222   0.714286   0.833333   \n",
       "recall      0.764706   0.944444   0.833333   0.722222   0.555556   0.833333   \n",
       "f1-score    0.764706   0.918919   0.810811   0.722222   0.625000   0.833333   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.875000   0.812500   0.590909   0.571429      0.75    0.756860   \n",
       "recall      0.777778   0.722222   0.722222   0.631579      0.75    0.750740   \n",
       "f1-score    0.823529   0.764706   0.650000   0.600000      0.75    0.751323   \n",
       "support    18.000000  18.000000  18.000000  19.000000      0.75  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.755786  \n",
       "recall         0.750000  \n",
       "f1-score       0.750408  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with all features\n",
    "parameters = {'kernel':['rbf'], 'C':[7.89,7.90,7.91,7.92,7.93]}\n",
    "svc=svm.SVC()\n",
    "clf2o = GridSearchCV(svc, parameters, cv=5, scoring='f1_macro')\n",
    "clf2o.fit(xo_train,yo_train)\n",
    "print(clf2o.best_score_,clf2o.best_params_)\n",
    "yo_pred=clf2o.predict(xo_test)\n",
    "report=pd.DataFrame(classification_report(yo_test,yo_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Random Forest Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7691224892253159 {'n_estimators': 100, 'n_jobs': -1, 'random_state': 42}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.631579</td>\n",
       "      <td>0.937500</td>\n",
       "      <td>0.684211</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>0.714286</td>\n",
       "      <td>0.842105</td>\n",
       "      <td>0.761905</td>\n",
       "      <td>0.647059</td>\n",
       "      <td>0.785714</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>0.766150</td>\n",
       "      <td>0.767006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.705882</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.578947</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>0.761816</td>\n",
       "      <td>0.761111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.882353</td>\n",
       "      <td>0.702703</td>\n",
       "      <td>0.842105</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.769231</td>\n",
       "      <td>0.864865</td>\n",
       "      <td>0.820513</td>\n",
       "      <td>0.628571</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>0.759367</td>\n",
       "      <td>0.759367</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.631579   0.937500   0.684211   0.800000   0.857143   0.714286   \n",
       "recall      0.705882   0.833333   0.722222   0.888889   0.666667   0.833333   \n",
       "f1-score    0.666667   0.882353   0.702703   0.842105   0.750000   0.769231   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.842105   0.761905   0.647059   0.785714  0.761111    0.766150   \n",
       "recall      0.888889   0.888889   0.611111   0.578947  0.761111    0.761816   \n",
       "f1-score    0.864865   0.820513   0.628571   0.666667  0.761111    0.759367   \n",
       "support    18.000000  18.000000  18.000000  19.000000  0.761111  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.767006  \n",
       "recall         0.761111  \n",
       "f1-score       0.759367  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with extracted features\n",
    "parameters = {'n_estimators':[99,100,101],'n_jobs':[-1], 'random_state':[42]}\n",
    "ran=RandomForestClassifier()\n",
    "clf3 = GridSearchCV(ran, parameters, cv=5, scoring='f1_macro')\n",
    "clf3.fit(x_train,y_train)\n",
    "print(clf3.best_score_,clf3.best_params_)\n",
    "y_pred=clf3.predict(x_test)\n",
    "report=pd.DataFrame(classification_report(y_test,y_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7647242998572188 {'n_estimators': 96, 'n_jobs': -1, 'random_state': 42}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.571429</td>\n",
       "      <td>0.941176</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.850000</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>0.823529</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.809524</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.755556</td>\n",
       "      <td>0.758540</td>\n",
       "      <td>0.758699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.470588</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.473684</td>\n",
       "      <td>0.755556</td>\n",
       "      <td>0.755538</td>\n",
       "      <td>0.755556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.516129</td>\n",
       "      <td>0.914286</td>\n",
       "      <td>0.697674</td>\n",
       "      <td>0.894737</td>\n",
       "      <td>0.774194</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.842105</td>\n",
       "      <td>0.871795</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.529412</td>\n",
       "      <td>0.755556</td>\n",
       "      <td>0.750700</td>\n",
       "      <td>0.750774</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.755556</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.571429   0.941176   0.600000   0.850000   0.923077   0.823529   \n",
       "recall      0.470588   0.888889   0.833333   0.944444   0.666667   0.777778   \n",
       "f1-score    0.516129   0.914286   0.697674   0.894737   0.774194   0.800000   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.800000   0.809524   0.666667   0.600000  0.755556    0.758540   \n",
       "recall      0.888889   0.944444   0.666667   0.473684  0.755556    0.755538   \n",
       "f1-score    0.842105   0.871795   0.666667   0.529412  0.755556    0.750700   \n",
       "support    18.000000  18.000000  18.000000  19.000000  0.755556  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.758699  \n",
       "recall         0.755556  \n",
       "f1-score       0.750774  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with all features\n",
    "parameters = {'n_estimators':[95,96,97,98,99],'n_jobs':[-1], 'random_state':[42]}\n",
    "ran=RandomForestClassifier()\n",
    "clf3o = GridSearchCV(ran, parameters, cv=5, scoring='f1_macro')\n",
    "clf3o.fit(xo_train,yo_train)\n",
    "print(clf3o.best_score_,clf3o.best_params_)\n",
    "yo_pred=clf3o.predict(xo_test)\n",
    "report=pd.DataFrame(classification_report(yo_test,yo_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Gaussian Naive Bayes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.4615699173930981 {}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.245283</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.375000</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.695652</td>\n",
       "      <td>0.454545</td>\n",
       "      <td>0.125000</td>\n",
       "      <td>0.455556</td>\n",
       "      <td>0.467604</td>\n",
       "      <td>0.466935</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.764706</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.055556</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.277778</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.277778</td>\n",
       "      <td>0.052632</td>\n",
       "      <td>0.455556</td>\n",
       "      <td>0.459512</td>\n",
       "      <td>0.455556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.371429</td>\n",
       "      <td>0.823529</td>\n",
       "      <td>0.090909</td>\n",
       "      <td>0.352941</td>\n",
       "      <td>0.370370</td>\n",
       "      <td>0.428571</td>\n",
       "      <td>0.625000</td>\n",
       "      <td>0.780488</td>\n",
       "      <td>0.344828</td>\n",
       "      <td>0.074074</td>\n",
       "      <td>0.455556</td>\n",
       "      <td>0.426214</td>\n",
       "      <td>0.424562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.455556</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.245283   0.875000   0.250000   0.375000   0.555556   0.600000   \n",
       "recall      0.764706   0.777778   0.055556   0.333333   0.277778   0.333333   \n",
       "f1-score    0.371429   0.823529   0.090909   0.352941   0.370370   0.428571   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.500000   0.695652   0.454545   0.125000  0.455556    0.467604   \n",
       "recall      0.833333   0.888889   0.277778   0.052632  0.455556    0.459512   \n",
       "f1-score    0.625000   0.780488   0.344828   0.074074  0.455556    0.426214   \n",
       "support    18.000000  18.000000  18.000000  19.000000  0.455556  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.466935  \n",
       "recall         0.455556  \n",
       "f1-score       0.424562  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# With selected features\n",
    "parameters = {}\n",
    "gnb=GaussianNB()\n",
    "clf4 = GridSearchCV(gnb, parameters, cv=5, scoring='f1_macro')\n",
    "clf4.fit(x_train,y_train)\n",
    "print(clf4.best_score_,clf4.best_params_)\n",
    "y_pred=clf4.predict(x_test)\n",
    "report=pd.DataFrame(classification_report(y_test,y_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5284561699135985 {}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.218750</td>\n",
       "      <td>0.937500</td>\n",
       "      <td>0.588235</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>0.428571</td>\n",
       "      <td>0.761905</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.125000</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.542663</td>\n",
       "      <td>0.542142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.411765</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.388889</td>\n",
       "      <td>0.277778</td>\n",
       "      <td>0.388889</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.388889</td>\n",
       "      <td>0.052632</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.501995</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.285714</td>\n",
       "      <td>0.882353</td>\n",
       "      <td>0.571429</td>\n",
       "      <td>0.358974</td>\n",
       "      <td>0.416667</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.566038</td>\n",
       "      <td>0.820513</td>\n",
       "      <td>0.437500</td>\n",
       "      <td>0.074074</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.491326</td>\n",
       "      <td>0.490150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.5</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.218750   0.937500   0.588235   0.333333   0.833333   0.700000   \n",
       "recall      0.411765   0.833333   0.555556   0.388889   0.277778   0.388889   \n",
       "f1-score    0.285714   0.882353   0.571429   0.358974   0.416667   0.500000   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.428571   0.761905   0.500000   0.125000       0.5    0.542663   \n",
       "recall      0.833333   0.888889   0.388889   0.052632       0.5    0.501995   \n",
       "f1-score    0.566038   0.820513   0.437500   0.074074       0.5    0.491326   \n",
       "support    18.000000  18.000000  18.000000  19.000000       0.5  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.542142  \n",
       "recall         0.500000  \n",
       "f1-score       0.490150  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with all features\n",
    "parameters = {}\n",
    "gnb=GaussianNB()\n",
    "clf4o = GridSearchCV(gnb, parameters, cv=5, scoring='f1_macro')\n",
    "clf4o.fit(xo_train,yo_train)\n",
    "print(clf4o.best_score_,clf4o.best_params_)\n",
    "yo_pred=clf4o.predict(xo_test)\n",
    "report=pd.DataFrame(classification_report(yo_test,yo_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Robust Scalar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "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>chroma_stft_mean</th>\n",
       "      <th>length</th>\n",
       "      <th>perceptr_var</th>\n",
       "      <th>rolloff_mean</th>\n",
       "      <th>spectral_centroid_mean</th>\n",
       "      <th>spectral_bandwidth_mean</th>\n",
       "      <th>spectral_centroid_var</th>\n",
       "      <th>mfcc1_mean</th>\n",
       "      <th>rms_mean</th>\n",
       "      <th>rms_var</th>\n",
       "      <th>...</th>\n",
       "      <th>mfcc4_var</th>\n",
       "      <th>mfcc9_mean</th>\n",
       "      <th>harmony_var</th>\n",
       "      <th>mfcc11_mean</th>\n",
       "      <th>mfcc3_mean</th>\n",
       "      <th>mfcc7_var</th>\n",
       "      <th>mfcc4_var</th>\n",
       "      <th>mfcc5_var</th>\n",
       "      <th>mfcc1_var</th>\n",
       "      <th>mfcc17_mean</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.403282</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>0.963965</td>\n",
       "      <td>0.885190</td>\n",
       "      <td>0.800547</td>\n",
       "      <td>1.150193</td>\n",
       "      <td>2.358739</td>\n",
       "      <td>0.191500</td>\n",
       "      <td>0.251539</td>\n",
       "      <td>1.121435</td>\n",
       "      <td>...</td>\n",
       "      <td>1.963770</td>\n",
       "      <td>0.109080</td>\n",
       "      <td>0.017521</td>\n",
       "      <td>0.646441</td>\n",
       "      <td>0.491454</td>\n",
       "      <td>1.186455</td>\n",
       "      <td>1.963770</td>\n",
       "      <td>1.905493</td>\n",
       "      <td>1.015182</td>\n",
       "      <td>-0.139137</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.110958</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-0.128376</td>\n",
       "      <td>-0.175230</td>\n",
       "      <td>-0.049460</td>\n",
       "      <td>-0.255211</td>\n",
       "      <td>-0.220711</td>\n",
       "      <td>-0.149113</td>\n",
       "      <td>-0.322669</td>\n",
       "      <td>-0.034318</td>\n",
       "      <td>...</td>\n",
       "      <td>0.403632</td>\n",
       "      <td>0.420639</td>\n",
       "      <td>-0.326769</td>\n",
       "      <td>1.039044</td>\n",
       "      <td>-0.855700</td>\n",
       "      <td>0.517928</td>\n",
       "      <td>0.403632</td>\n",
       "      <td>1.603019</td>\n",
       "      <td>-0.147949</td>\n",
       "      <td>-0.605911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.085328</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.070860</td>\n",
       "      <td>-0.047066</td>\n",
       "      <td>0.090493</td>\n",
       "      <td>0.495253</td>\n",
       "      <td>0.497466</td>\n",
       "      <td>-0.011588</td>\n",
       "      <td>-0.234839</td>\n",
       "      <td>-0.250317</td>\n",
       "      <td>...</td>\n",
       "      <td>0.638859</td>\n",
       "      <td>-0.555567</td>\n",
       "      <td>-0.295017</td>\n",
       "      <td>0.373312</td>\n",
       "      <td>0.147288</td>\n",
       "      <td>-0.336138</td>\n",
       "      <td>0.638859</td>\n",
       "      <td>0.581212</td>\n",
       "      <td>-0.451297</td>\n",
       "      <td>0.086414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.330723</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.104082</td>\n",
       "      <td>-0.398648</td>\n",
       "      <td>-0.311733</td>\n",
       "      <td>-0.471100</td>\n",
       "      <td>-0.205331</td>\n",
       "      <td>0.203181</td>\n",
       "      <td>0.625759</td>\n",
       "      <td>0.669612</td>\n",
       "      <td>...</td>\n",
       "      <td>0.847088</td>\n",
       "      <td>-0.053313</td>\n",
       "      <td>0.441084</td>\n",
       "      <td>0.040488</td>\n",
       "      <td>-0.846263</td>\n",
       "      <td>0.625288</td>\n",
       "      <td>0.847088</td>\n",
       "      <td>1.179218</td>\n",
       "      <td>0.013667</td>\n",
       "      <td>-0.385017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.603013</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.149218</td>\n",
       "      <td>-0.231222</td>\n",
       "      <td>0.041714</td>\n",
       "      <td>-0.447105</td>\n",
       "      <td>-0.490331</td>\n",
       "      <td>0.143605</td>\n",
       "      <td>-0.261650</td>\n",
       "      <td>-0.547811</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.536810</td>\n",
       "      <td>-0.897523</td>\n",
       "      <td>-0.363837</td>\n",
       "      <td>-0.782861</td>\n",
       "      <td>-1.605150</td>\n",
       "      <td>0.080621</td>\n",
       "      <td>-0.536810</td>\n",
       "      <td>-0.468728</td>\n",
       "      <td>-0.561952</td>\n",
       "      <td>-1.355155</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",
       "      <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>895</th>\n",
       "      <td>0.720897</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.012372</td>\n",
       "      <td>0.521569</td>\n",
       "      <td>0.595786</td>\n",
       "      <td>0.695232</td>\n",
       "      <td>2.461122</td>\n",
       "      <td>-0.064993</td>\n",
       "      <td>0.266627</td>\n",
       "      <td>2.453950</td>\n",
       "      <td>...</td>\n",
       "      <td>0.439862</td>\n",
       "      <td>0.942340</td>\n",
       "      <td>0.384982</td>\n",
       "      <td>1.178448</td>\n",
       "      <td>0.505446</td>\n",
       "      <td>0.584584</td>\n",
       "      <td>0.439862</td>\n",
       "      <td>0.270910</td>\n",
       "      <td>0.591393</td>\n",
       "      <td>-0.229144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>896</th>\n",
       "      <td>-0.813590</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.279710</td>\n",
       "      <td>-1.139529</td>\n",
       "      <td>-1.030885</td>\n",
       "      <td>-1.159159</td>\n",
       "      <td>-0.483289</td>\n",
       "      <td>-1.011355</td>\n",
       "      <td>-0.356706</td>\n",
       "      <td>0.496477</td>\n",
       "      <td>...</td>\n",
       "      <td>0.639573</td>\n",
       "      <td>0.694050</td>\n",
       "      <td>-0.161506</td>\n",
       "      <td>0.513291</td>\n",
       "      <td>0.376565</td>\n",
       "      <td>-0.113555</td>\n",
       "      <td>0.639573</td>\n",
       "      <td>0.243463</td>\n",
       "      <td>0.284403</td>\n",
       "      <td>0.207094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>897</th>\n",
       "      <td>0.591098</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>3.452005</td>\n",
       "      <td>0.904605</td>\n",
       "      <td>0.831542</td>\n",
       "      <td>1.144908</td>\n",
       "      <td>1.475929</td>\n",
       "      <td>0.590582</td>\n",
       "      <td>1.164727</td>\n",
       "      <td>4.272730</td>\n",
       "      <td>...</td>\n",
       "      <td>0.936927</td>\n",
       "      <td>0.832362</td>\n",
       "      <td>0.799935</td>\n",
       "      <td>0.129619</td>\n",
       "      <td>0.516824</td>\n",
       "      <td>1.315846</td>\n",
       "      <td>0.936927</td>\n",
       "      <td>0.974422</td>\n",
       "      <td>0.484489</td>\n",
       "      <td>0.359131</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>898</th>\n",
       "      <td>0.446919</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.065750</td>\n",
       "      <td>-0.316803</td>\n",
       "      <td>-0.372249</td>\n",
       "      <td>0.459685</td>\n",
       "      <td>0.986822</td>\n",
       "      <td>0.006570</td>\n",
       "      <td>1.496732</td>\n",
       "      <td>4.503363</td>\n",
       "      <td>...</td>\n",
       "      <td>0.531397</td>\n",
       "      <td>1.110681</td>\n",
       "      <td>2.680063</td>\n",
       "      <td>1.482719</td>\n",
       "      <td>1.063745</td>\n",
       "      <td>0.046170</td>\n",
       "      <td>0.531397</td>\n",
       "      <td>0.793820</td>\n",
       "      <td>0.433566</td>\n",
       "      <td>-0.212235</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>899</th>\n",
       "      <td>0.872529</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.067789</td>\n",
       "      <td>0.255981</td>\n",
       "      <td>0.457408</td>\n",
       "      <td>0.076421</td>\n",
       "      <td>-0.226553</td>\n",
       "      <td>0.728724</td>\n",
       "      <td>0.979310</td>\n",
       "      <td>-0.131821</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.188437</td>\n",
       "      <td>-0.305962</td>\n",
       "      <td>1.035382</td>\n",
       "      <td>-0.821721</td>\n",
       "      <td>-0.685494</td>\n",
       "      <td>0.059087</td>\n",
       "      <td>-0.188437</td>\n",
       "      <td>-0.507880</td>\n",
       "      <td>-0.503302</td>\n",
       "      <td>-0.396394</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>900 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     chroma_stft_mean  length  perceptr_var  rolloff_mean  \\\n",
       "0            0.403282    -1.0      0.963965      0.885190   \n",
       "1            0.110958    -1.0     -0.128376     -0.175230   \n",
       "2           -0.085328     0.0     -0.070860     -0.047066   \n",
       "3            0.330723     0.0      1.104082     -0.398648   \n",
       "4            0.603013     0.0     -0.149218     -0.231222   \n",
       "..                ...     ...           ...           ...   \n",
       "895          0.720897    -1.0      1.012372      0.521569   \n",
       "896         -0.813590     0.0     -0.279710     -1.139529   \n",
       "897          0.591098    -1.0      3.452005      0.904605   \n",
       "898          0.446919     0.0      2.065750     -0.316803   \n",
       "899          0.872529     0.0      1.067789      0.255981   \n",
       "\n",
       "     spectral_centroid_mean  spectral_bandwidth_mean  spectral_centroid_var  \\\n",
       "0                  0.800547                 1.150193               2.358739   \n",
       "1                 -0.049460                -0.255211              -0.220711   \n",
       "2                  0.090493                 0.495253               0.497466   \n",
       "3                 -0.311733                -0.471100              -0.205331   \n",
       "4                  0.041714                -0.447105              -0.490331   \n",
       "..                      ...                      ...                    ...   \n",
       "895                0.595786                 0.695232               2.461122   \n",
       "896               -1.030885                -1.159159              -0.483289   \n",
       "897                0.831542                 1.144908               1.475929   \n",
       "898               -0.372249                 0.459685               0.986822   \n",
       "899                0.457408                 0.076421              -0.226553   \n",
       "\n",
       "     mfcc1_mean  rms_mean   rms_var  ...  mfcc4_var  mfcc9_mean  harmony_var  \\\n",
       "0      0.191500  0.251539  1.121435  ...   1.963770    0.109080     0.017521   \n",
       "1     -0.149113 -0.322669 -0.034318  ...   0.403632    0.420639    -0.326769   \n",
       "2     -0.011588 -0.234839 -0.250317  ...   0.638859   -0.555567    -0.295017   \n",
       "3      0.203181  0.625759  0.669612  ...   0.847088   -0.053313     0.441084   \n",
       "4      0.143605 -0.261650 -0.547811  ...  -0.536810   -0.897523    -0.363837   \n",
       "..          ...       ...       ...  ...        ...         ...          ...   \n",
       "895   -0.064993  0.266627  2.453950  ...   0.439862    0.942340     0.384982   \n",
       "896   -1.011355 -0.356706  0.496477  ...   0.639573    0.694050    -0.161506   \n",
       "897    0.590582  1.164727  4.272730  ...   0.936927    0.832362     0.799935   \n",
       "898    0.006570  1.496732  4.503363  ...   0.531397    1.110681     2.680063   \n",
       "899    0.728724  0.979310 -0.131821  ...  -0.188437   -0.305962     1.035382   \n",
       "\n",
       "     mfcc11_mean  mfcc3_mean  mfcc7_var  mfcc4_var  mfcc5_var  mfcc1_var  \\\n",
       "0       0.646441    0.491454   1.186455   1.963770   1.905493   1.015182   \n",
       "1       1.039044   -0.855700   0.517928   0.403632   1.603019  -0.147949   \n",
       "2       0.373312    0.147288  -0.336138   0.638859   0.581212  -0.451297   \n",
       "3       0.040488   -0.846263   0.625288   0.847088   1.179218   0.013667   \n",
       "4      -0.782861   -1.605150   0.080621  -0.536810  -0.468728  -0.561952   \n",
       "..           ...         ...        ...        ...        ...        ...   \n",
       "895     1.178448    0.505446   0.584584   0.439862   0.270910   0.591393   \n",
       "896     0.513291    0.376565  -0.113555   0.639573   0.243463   0.284403   \n",
       "897     0.129619    0.516824   1.315846   0.936927   0.974422   0.484489   \n",
       "898     1.482719    1.063745   0.046170   0.531397   0.793820   0.433566   \n",
       "899    -0.821721   -0.685494   0.059087  -0.188437  -0.507880  -0.503302   \n",
       "\n",
       "     mfcc17_mean  \n",
       "0      -0.139137  \n",
       "1      -0.605911  \n",
       "2       0.086414  \n",
       "3      -0.385017  \n",
       "4      -1.355155  \n",
       "..           ...  \n",
       "895    -0.229144  \n",
       "896     0.207094  \n",
       "897     0.359131  \n",
       "898    -0.212235  \n",
       "899    -0.396394  \n",
       "\n",
       "[900 rows x 32 columns]"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler = RobustScaler()\n",
    "scaler.fit(xc)\n",
    "xcr=pd.DataFrame(scaler.transform(xc),columns=xc.columns)\n",
    "scaler.fit(x)\n",
    "xr=pd.DataFrame(scaler.transform(x),columns=x.columns)\n",
    "xcr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "xr_train, xr_test, yr_train, yr_test = train_test_split(xcr, yc, test_size=0.20, random_state=42, stratify=yc)\n",
    "xro_train, xro_test, yro_train, yro_test = train_test_split(xr, yc, test_size=0.20, random_state=42, stratify=yc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Support Vector Machine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.769581224757198 {'C': 7.99, 'kernel': 'rbf'}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.785714</td>\n",
       "      <td>0.933333</td>\n",
       "      <td>0.714286</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.625000</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>0.775726</td>\n",
       "      <td>0.775250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.647059</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.736842</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>0.760612</td>\n",
       "      <td>0.761111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.709677</td>\n",
       "      <td>0.848485</td>\n",
       "      <td>0.769231</td>\n",
       "      <td>0.717949</td>\n",
       "      <td>0.709677</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.823529</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.714286</td>\n",
       "      <td>0.717949</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>0.762189</td>\n",
       "      <td>0.762235</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.785714   0.933333   0.714286   0.666667   0.846154   0.777778   \n",
       "recall      0.647059   0.777778   0.833333   0.777778   0.611111   0.777778   \n",
       "f1-score    0.709677   0.848485   0.769231   0.717949   0.709677   0.777778   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.875000   0.833333   0.625000   0.700000  0.761111    0.775726   \n",
       "recall      0.777778   0.833333   0.833333   0.736842  0.761111    0.760612   \n",
       "f1-score    0.823529   0.833333   0.714286   0.717949  0.761111    0.762189   \n",
       "support    18.000000  18.000000  18.000000  19.000000  0.761111  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.775250  \n",
       "recall         0.761111  \n",
       "f1-score       0.762235  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with extracted features\n",
    "parameters = {'kernel':['rbf'], 'C':[7.96,7.97,7.98,7.99,8]}\n",
    "svc=svm.SVC()\n",
    "clf2a = GridSearchCV(svc, parameters, cv=5, scoring='f1_macro')\n",
    "clf2a.fit(xr_train,yr_train)\n",
    "print(clf2a.best_score_,clf2a.best_params_)\n",
    "yr_pred=clf2a.predict(xr_test)\n",
    "report=pd.DataFrame(classification_report(yr_test,yr_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7231076273375014 {'C': 7.96, 'kernel': 'rbf'}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.692308</td>\n",
       "      <td>0.937500</td>\n",
       "      <td>0.560000</td>\n",
       "      <td>0.526316</td>\n",
       "      <td>0.916667</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>0.684211</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>0.608696</td>\n",
       "      <td>0.619048</td>\n",
       "      <td>0.705556</td>\n",
       "      <td>0.729474</td>\n",
       "      <td>0.729067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.529412</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.684211</td>\n",
       "      <td>0.705556</td>\n",
       "      <td>0.704696</td>\n",
       "      <td>0.705556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.882353</td>\n",
       "      <td>0.651163</td>\n",
       "      <td>0.540541</td>\n",
       "      <td>0.733333</td>\n",
       "      <td>0.823529</td>\n",
       "      <td>0.702703</td>\n",
       "      <td>0.823529</td>\n",
       "      <td>0.682927</td>\n",
       "      <td>0.650000</td>\n",
       "      <td>0.705556</td>\n",
       "      <td>0.709008</td>\n",
       "      <td>0.709286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.705556</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.692308   0.937500   0.560000   0.526316   0.916667   0.875000   \n",
       "recall      0.529412   0.833333   0.777778   0.555556   0.611111   0.777778   \n",
       "f1-score    0.600000   0.882353   0.651163   0.540541   0.733333   0.823529   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.684211   0.875000   0.608696   0.619048  0.705556    0.729474   \n",
       "recall      0.722222   0.777778   0.777778   0.684211  0.705556    0.704696   \n",
       "f1-score    0.702703   0.823529   0.682927   0.650000  0.705556    0.709008   \n",
       "support    18.000000  18.000000  18.000000  19.000000  0.705556  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.729067  \n",
       "recall         0.705556  \n",
       "f1-score       0.709286  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with all features\n",
    "parameters = {'kernel':['rbf'], 'C':[7.96,7.97,7.98,7.99,8]}\n",
    "svc=svm.SVC()\n",
    "clf2ao = GridSearchCV(svc, parameters, cv=5, scoring='f1_macro')\n",
    "clf2ao.fit(xro_train,yro_train)\n",
    "print(clf2ao.best_score_,clf2ao.best_params_)\n",
    "yro_pred=clf2ao.predict(xro_test)\n",
    "report=pd.DataFrame(classification_report(yro_test,yro_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Random Forest Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7691224892253159 {'n_estimators': 100, 'n_jobs': -1, 'random_state': 42}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.631579</td>\n",
       "      <td>0.937500</td>\n",
       "      <td>0.684211</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>0.714286</td>\n",
       "      <td>0.842105</td>\n",
       "      <td>0.761905</td>\n",
       "      <td>0.647059</td>\n",
       "      <td>0.785714</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>0.766150</td>\n",
       "      <td>0.767006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.705882</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.578947</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>0.761816</td>\n",
       "      <td>0.761111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.882353</td>\n",
       "      <td>0.702703</td>\n",
       "      <td>0.842105</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.769231</td>\n",
       "      <td>0.864865</td>\n",
       "      <td>0.820513</td>\n",
       "      <td>0.628571</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>0.759367</td>\n",
       "      <td>0.759367</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.631579   0.937500   0.684211   0.800000   0.857143   0.714286   \n",
       "recall      0.705882   0.833333   0.722222   0.888889   0.666667   0.833333   \n",
       "f1-score    0.666667   0.882353   0.702703   0.842105   0.750000   0.769231   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.842105   0.761905   0.647059   0.785714  0.761111    0.766150   \n",
       "recall      0.888889   0.888889   0.611111   0.578947  0.761111    0.761816   \n",
       "f1-score    0.864865   0.820513   0.628571   0.666667  0.761111    0.759367   \n",
       "support    18.000000  18.000000  18.000000  19.000000  0.761111  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.767006  \n",
       "recall         0.761111  \n",
       "f1-score       0.759367  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with extracted features\n",
    "parameters = {'n_estimators':[99,100,101],'n_jobs':[-1], 'random_state':[42]}\n",
    "ran=RandomForestClassifier()\n",
    "clf3 = GridSearchCV(ran, parameters, cv=5, scoring='f1_macro')\n",
    "clf3.fit(xr_train,yr_train)\n",
    "print(clf3.best_score_,clf3.best_params_)\n",
    "yr_pred=clf3.predict(xr_test)\n",
    "report=pd.DataFrame(classification_report(yr_test,yr_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7638448767569391 {'n_estimators': 99, 'n_jobs': -1, 'random_state': 42}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.533333</td>\n",
       "      <td>0.941176</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>0.809524</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>0.823529</td>\n",
       "      <td>0.789474</td>\n",
       "      <td>0.809524</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.744444</td>\n",
       "      <td>0.747964</td>\n",
       "      <td>0.748334</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.470588</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.473684</td>\n",
       "      <td>0.744444</td>\n",
       "      <td>0.744427</td>\n",
       "      <td>0.744444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.914286</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.871795</td>\n",
       "      <td>0.774194</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.810811</td>\n",
       "      <td>0.871795</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.529412</td>\n",
       "      <td>0.744444</td>\n",
       "      <td>0.740562</td>\n",
       "      <td>0.740726</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.744444</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.533333   0.941176   0.583333   0.809524   0.923077   0.823529   \n",
       "recall      0.470588   0.888889   0.777778   0.944444   0.666667   0.777778   \n",
       "f1-score    0.500000   0.914286   0.666667   0.871795   0.774194   0.800000   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.789474   0.809524   0.666667   0.600000  0.744444    0.747964   \n",
       "recall      0.833333   0.944444   0.666667   0.473684  0.744444    0.744427   \n",
       "f1-score    0.810811   0.871795   0.666667   0.529412  0.744444    0.740562   \n",
       "support    18.000000  18.000000  18.000000  19.000000  0.744444  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.748334  \n",
       "recall         0.744444  \n",
       "f1-score       0.740726  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with all features\n",
    "parameters = {'n_estimators':[97,98,99,100,101],'n_jobs':[-1], 'random_state':[42]}\n",
    "ran=RandomForestClassifier()\n",
    "clf3o = GridSearchCV(ran, parameters, cv=5, scoring='f1_macro')\n",
    "clf3o.fit(xro_train,yro_train)\n",
    "print(clf3o.best_score_,clf3o.best_params_)\n",
    "yro_pred=clf3o.predict(xro_test)\n",
    "report=pd.DataFrame(classification_report(yro_test,yro_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Bagging Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6813229764191518 {'n_estimators': 14, 'n_jobs': -1, 'random_state': 42}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.666667</td>\n",
       "      <td>0.761905</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>0.818182</td>\n",
       "      <td>0.785714</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>0.789474</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.636364</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.749164</td>\n",
       "      <td>0.749164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.470588</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.526316</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.749690</td>\n",
       "      <td>0.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.551724</td>\n",
       "      <td>0.820513</td>\n",
       "      <td>0.736842</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.687500</td>\n",
       "      <td>0.787879</td>\n",
       "      <td>0.810811</td>\n",
       "      <td>0.842105</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>0.588235</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.742561</td>\n",
       "      <td>0.742764</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.75</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.666667   0.761905   0.700000   0.818182   0.785714   0.866667   \n",
       "recall      0.470588   0.888889   0.777778   1.000000   0.611111   0.722222   \n",
       "f1-score    0.551724   0.820513   0.736842   0.900000   0.687500   0.787879   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.789474   0.800000   0.636364   0.666667      0.75    0.749164   \n",
       "recall      0.833333   0.888889   0.777778   0.526316      0.75    0.749690   \n",
       "f1-score    0.810811   0.842105   0.700000   0.588235      0.75    0.742561   \n",
       "support    18.000000  18.000000  18.000000  19.000000      0.75  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.749164  \n",
       "recall         0.750000  \n",
       "f1-score       0.742764  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with selected features\n",
    "parameters = {'n_estimators':[13,14,15],'n_jobs':[-1], 'random_state':[42]}\n",
    "bag=BaggingClassifier()\n",
    "clf5 = GridSearchCV(bag, parameters, cv=5, scoring='f1_macro')\n",
    "clf5.fit(xr_train,yr_train)\n",
    "print(clf5.best_score_,clf5.best_params_)\n",
    "yr_pred=clf5.predict(xr_test)\n",
    "report=pd.DataFrame(classification_report(yr_test,yr_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7121198719429203 {'n_estimators': 15, 'n_jobs': -1, 'random_state': 42}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.600000</td>\n",
       "      <td>0.842105</td>\n",
       "      <td>0.520000</td>\n",
       "      <td>0.850000</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>0.916667</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.739130</td>\n",
       "      <td>0.619048</td>\n",
       "      <td>0.642857</td>\n",
       "      <td>0.727778</td>\n",
       "      <td>0.740929</td>\n",
       "      <td>0.741168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.529412</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.473684</td>\n",
       "      <td>0.727778</td>\n",
       "      <td>0.728087</td>\n",
       "      <td>0.727778</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.562500</td>\n",
       "      <td>0.864865</td>\n",
       "      <td>0.604651</td>\n",
       "      <td>0.894737</td>\n",
       "      <td>0.709677</td>\n",
       "      <td>0.733333</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.829268</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.545455</td>\n",
       "      <td>0.727778</td>\n",
       "      <td>0.724449</td>\n",
       "      <td>0.724354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.727778</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.600000   0.842105   0.520000   0.850000   0.846154   0.916667   \n",
       "recall      0.529412   0.888889   0.722222   0.944444   0.611111   0.611111   \n",
       "f1-score    0.562500   0.864865   0.604651   0.894737   0.709677   0.733333   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.833333   0.739130   0.619048   0.642857  0.727778    0.740929   \n",
       "recall      0.833333   0.944444   0.722222   0.473684  0.727778    0.728087   \n",
       "f1-score    0.833333   0.829268   0.666667   0.545455  0.727778    0.724449   \n",
       "support    18.000000  18.000000  18.000000  19.000000  0.727778  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.741168  \n",
       "recall         0.727778  \n",
       "f1-score       0.724354  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with all features\n",
    "parameters = {'n_estimators':[14,15,16],'n_jobs':[-1], 'random_state':[42]}\n",
    "bag=BaggingClassifier()\n",
    "clf5o = GridSearchCV(bag, parameters, cv=5, scoring='f1_macro')\n",
    "clf5o.fit(xro_train,yro_train)\n",
    "print(clf5o.best_score_,clf5o.best_params_)\n",
    "yro_pred=clf5o.predict(xro_test)\n",
    "report=pd.DataFrame(classification_report(yro_test,yro_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Minmax Scalar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "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>chroma_stft_mean</th>\n",
       "      <th>length</th>\n",
       "      <th>perceptr_var</th>\n",
       "      <th>rolloff_mean</th>\n",
       "      <th>spectral_centroid_mean</th>\n",
       "      <th>spectral_bandwidth_mean</th>\n",
       "      <th>spectral_centroid_var</th>\n",
       "      <th>mfcc1_mean</th>\n",
       "      <th>rms_mean</th>\n",
       "      <th>rms_var</th>\n",
       "      <th>...</th>\n",
       "      <th>mfcc4_var</th>\n",
       "      <th>mfcc9_mean</th>\n",
       "      <th>harmony_var</th>\n",
       "      <th>mfcc11_mean</th>\n",
       "      <th>mfcc3_mean</th>\n",
       "      <th>mfcc7_var</th>\n",
       "      <th>mfcc4_var</th>\n",
       "      <th>mfcc5_var</th>\n",
       "      <th>mfcc1_var</th>\n",
       "      <th>mfcc17_mean</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.530572</td>\n",
       "      <td>0.095142</td>\n",
       "      <td>0.200515</td>\n",
       "      <td>0.734768</td>\n",
       "      <td>0.651867</td>\n",
       "      <td>0.794782</td>\n",
       "      <td>0.430339</td>\n",
       "      <td>0.770686</td>\n",
       "      <td>0.348285</td>\n",
       "      <td>0.171782</td>\n",
       "      <td>...</td>\n",
       "      <td>0.742328</td>\n",
       "      <td>0.502755</td>\n",
       "      <td>0.104326</td>\n",
       "      <td>0.624233</td>\n",
       "      <td>0.640720</td>\n",
       "      <td>0.427275</td>\n",
       "      <td>0.742328</td>\n",
       "      <td>0.344795</td>\n",
       "      <td>0.217235</td>\n",
       "      <td>0.425846</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.461259</td>\n",
       "      <td>0.095142</td>\n",
       "      <td>0.056068</td>\n",
       "      <td>0.448646</td>\n",
       "      <td>0.414080</td>\n",
       "      <td>0.442797</td>\n",
       "      <td>0.078950</td>\n",
       "      <td>0.696779</td>\n",
       "      <td>0.218712</td>\n",
       "      <td>0.061847</td>\n",
       "      <td>...</td>\n",
       "      <td>0.356513</td>\n",
       "      <td>0.581706</td>\n",
       "      <td>0.054096</td>\n",
       "      <td>0.709421</td>\n",
       "      <td>0.372059</td>\n",
       "      <td>0.300131</td>\n",
       "      <td>0.356513</td>\n",
       "      <td>0.308558</td>\n",
       "      <td>0.087277</td>\n",
       "      <td>0.323413</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.414717</td>\n",
       "      <td>0.113487</td>\n",
       "      <td>0.063673</td>\n",
       "      <td>0.483227</td>\n",
       "      <td>0.453231</td>\n",
       "      <td>0.630752</td>\n",
       "      <td>0.176784</td>\n",
       "      <td>0.726620</td>\n",
       "      <td>0.238531</td>\n",
       "      <td>0.041302</td>\n",
       "      <td>...</td>\n",
       "      <td>0.414684</td>\n",
       "      <td>0.334327</td>\n",
       "      <td>0.058729</td>\n",
       "      <td>0.564969</td>\n",
       "      <td>0.572084</td>\n",
       "      <td>0.137702</td>\n",
       "      <td>0.414684</td>\n",
       "      <td>0.186141</td>\n",
       "      <td>0.053383</td>\n",
       "      <td>0.475342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.513368</td>\n",
       "      <td>0.113487</td>\n",
       "      <td>0.219044</td>\n",
       "      <td>0.388363</td>\n",
       "      <td>0.340710</td>\n",
       "      <td>0.388727</td>\n",
       "      <td>0.081045</td>\n",
       "      <td>0.773221</td>\n",
       "      <td>0.432729</td>\n",
       "      <td>0.128805</td>\n",
       "      <td>...</td>\n",
       "      <td>0.466178</td>\n",
       "      <td>0.461603</td>\n",
       "      <td>0.166122</td>\n",
       "      <td>0.492752</td>\n",
       "      <td>0.373941</td>\n",
       "      <td>0.320550</td>\n",
       "      <td>0.466178</td>\n",
       "      <td>0.257785</td>\n",
       "      <td>0.105334</td>\n",
       "      <td>0.371888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.577931</td>\n",
       "      <td>0.113487</td>\n",
       "      <td>0.053311</td>\n",
       "      <td>0.433538</td>\n",
       "      <td>0.439586</td>\n",
       "      <td>0.394737</td>\n",
       "      <td>0.042220</td>\n",
       "      <td>0.760294</td>\n",
       "      <td>0.232481</td>\n",
       "      <td>0.013004</td>\n",
       "      <td>...</td>\n",
       "      <td>0.123946</td>\n",
       "      <td>0.247673</td>\n",
       "      <td>0.048688</td>\n",
       "      <td>0.314100</td>\n",
       "      <td>0.222598</td>\n",
       "      <td>0.216963</td>\n",
       "      <td>0.123946</td>\n",
       "      <td>0.060355</td>\n",
       "      <td>0.041020</td>\n",
       "      <td>0.158994</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",
       "      <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>895</th>\n",
       "      <td>0.605883</td>\n",
       "      <td>0.095142</td>\n",
       "      <td>0.206916</td>\n",
       "      <td>0.636656</td>\n",
       "      <td>0.594585</td>\n",
       "      <td>0.680837</td>\n",
       "      <td>0.444286</td>\n",
       "      <td>0.715032</td>\n",
       "      <td>0.351690</td>\n",
       "      <td>0.298530</td>\n",
       "      <td>...</td>\n",
       "      <td>0.365473</td>\n",
       "      <td>0.713910</td>\n",
       "      <td>0.157937</td>\n",
       "      <td>0.739669</td>\n",
       "      <td>0.643511</td>\n",
       "      <td>0.312808</td>\n",
       "      <td>0.365473</td>\n",
       "      <td>0.148966</td>\n",
       "      <td>0.169884</td>\n",
       "      <td>0.406094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>896</th>\n",
       "      <td>0.242037</td>\n",
       "      <td>0.113487</td>\n",
       "      <td>0.036056</td>\n",
       "      <td>0.188459</td>\n",
       "      <td>0.139530</td>\n",
       "      <td>0.216402</td>\n",
       "      <td>0.043180</td>\n",
       "      <td>0.509687</td>\n",
       "      <td>0.211031</td>\n",
       "      <td>0.112336</td>\n",
       "      <td>...</td>\n",
       "      <td>0.414860</td>\n",
       "      <td>0.650991</td>\n",
       "      <td>0.078207</td>\n",
       "      <td>0.595342</td>\n",
       "      <td>0.617808</td>\n",
       "      <td>0.180033</td>\n",
       "      <td>0.414860</td>\n",
       "      <td>0.145678</td>\n",
       "      <td>0.135584</td>\n",
       "      <td>0.501825</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>897</th>\n",
       "      <td>0.575106</td>\n",
       "      <td>0.095142</td>\n",
       "      <td>0.529525</td>\n",
       "      <td>0.740006</td>\n",
       "      <td>0.660537</td>\n",
       "      <td>0.793459</td>\n",
       "      <td>0.310077</td>\n",
       "      <td>0.857281</td>\n",
       "      <td>0.554350</td>\n",
       "      <td>0.471531</td>\n",
       "      <td>...</td>\n",
       "      <td>0.488395</td>\n",
       "      <td>0.686040</td>\n",
       "      <td>0.218476</td>\n",
       "      <td>0.512092</td>\n",
       "      <td>0.645780</td>\n",
       "      <td>0.451883</td>\n",
       "      <td>0.488395</td>\n",
       "      <td>0.233249</td>\n",
       "      <td>0.157940</td>\n",
       "      <td>0.535189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>898</th>\n",
       "      <td>0.540919</td>\n",
       "      <td>0.113487</td>\n",
       "      <td>0.346211</td>\n",
       "      <td>0.410446</td>\n",
       "      <td>0.323781</td>\n",
       "      <td>0.621844</td>\n",
       "      <td>0.243448</td>\n",
       "      <td>0.730560</td>\n",
       "      <td>0.629269</td>\n",
       "      <td>0.493469</td>\n",
       "      <td>...</td>\n",
       "      <td>0.388109</td>\n",
       "      <td>0.756569</td>\n",
       "      <td>0.492778</td>\n",
       "      <td>0.805690</td>\n",
       "      <td>0.754852</td>\n",
       "      <td>0.210411</td>\n",
       "      <td>0.388109</td>\n",
       "      <td>0.211613</td>\n",
       "      <td>0.152250</td>\n",
       "      <td>0.409805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>899</th>\n",
       "      <td>0.641837</td>\n",
       "      <td>0.113487</td>\n",
       "      <td>0.214244</td>\n",
       "      <td>0.564995</td>\n",
       "      <td>0.555875</td>\n",
       "      <td>0.525855</td>\n",
       "      <td>0.078154</td>\n",
       "      <td>0.887255</td>\n",
       "      <td>0.512510</td>\n",
       "      <td>0.052573</td>\n",
       "      <td>...</td>\n",
       "      <td>0.210097</td>\n",
       "      <td>0.397579</td>\n",
       "      <td>0.252827</td>\n",
       "      <td>0.305668</td>\n",
       "      <td>0.406003</td>\n",
       "      <td>0.212867</td>\n",
       "      <td>0.210097</td>\n",
       "      <td>0.055664</td>\n",
       "      <td>0.047573</td>\n",
       "      <td>0.369391</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>900 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     chroma_stft_mean    length  perceptr_var  rolloff_mean  \\\n",
       "0            0.530572  0.095142      0.200515      0.734768   \n",
       "1            0.461259  0.095142      0.056068      0.448646   \n",
       "2            0.414717  0.113487      0.063673      0.483227   \n",
       "3            0.513368  0.113487      0.219044      0.388363   \n",
       "4            0.577931  0.113487      0.053311      0.433538   \n",
       "..                ...       ...           ...           ...   \n",
       "895          0.605883  0.095142      0.206916      0.636656   \n",
       "896          0.242037  0.113487      0.036056      0.188459   \n",
       "897          0.575106  0.095142      0.529525      0.740006   \n",
       "898          0.540919  0.113487      0.346211      0.410446   \n",
       "899          0.641837  0.113487      0.214244      0.564995   \n",
       "\n",
       "     spectral_centroid_mean  spectral_bandwidth_mean  spectral_centroid_var  \\\n",
       "0                  0.651867                 0.794782               0.430339   \n",
       "1                  0.414080                 0.442797               0.078950   \n",
       "2                  0.453231                 0.630752               0.176784   \n",
       "3                  0.340710                 0.388727               0.081045   \n",
       "4                  0.439586                 0.394737               0.042220   \n",
       "..                      ...                      ...                    ...   \n",
       "895                0.594585                 0.680837               0.444286   \n",
       "896                0.139530                 0.216402               0.043180   \n",
       "897                0.660537                 0.793459               0.310077   \n",
       "898                0.323781                 0.621844               0.243448   \n",
       "899                0.555875                 0.525855               0.078154   \n",
       "\n",
       "     mfcc1_mean  rms_mean   rms_var  ...  mfcc4_var  mfcc9_mean  harmony_var  \\\n",
       "0      0.770686  0.348285  0.171782  ...   0.742328    0.502755     0.104326   \n",
       "1      0.696779  0.218712  0.061847  ...   0.356513    0.581706     0.054096   \n",
       "2      0.726620  0.238531  0.041302  ...   0.414684    0.334327     0.058729   \n",
       "3      0.773221  0.432729  0.128805  ...   0.466178    0.461603     0.166122   \n",
       "4      0.760294  0.232481  0.013004  ...   0.123946    0.247673     0.048688   \n",
       "..          ...       ...       ...  ...        ...         ...          ...   \n",
       "895    0.715032  0.351690  0.298530  ...   0.365473    0.713910     0.157937   \n",
       "896    0.509687  0.211031  0.112336  ...   0.414860    0.650991     0.078207   \n",
       "897    0.857281  0.554350  0.471531  ...   0.488395    0.686040     0.218476   \n",
       "898    0.730560  0.629269  0.493469  ...   0.388109    0.756569     0.492778   \n",
       "899    0.887255  0.512510  0.052573  ...   0.210097    0.397579     0.252827   \n",
       "\n",
       "     mfcc11_mean  mfcc3_mean  mfcc7_var  mfcc4_var  mfcc5_var  mfcc1_var  \\\n",
       "0       0.624233    0.640720   0.427275   0.742328   0.344795   0.217235   \n",
       "1       0.709421    0.372059   0.300131   0.356513   0.308558   0.087277   \n",
       "2       0.564969    0.572084   0.137702   0.414684   0.186141   0.053383   \n",
       "3       0.492752    0.373941   0.320550   0.466178   0.257785   0.105334   \n",
       "4       0.314100    0.222598   0.216963   0.123946   0.060355   0.041020   \n",
       "..           ...         ...        ...        ...        ...        ...   \n",
       "895     0.739669    0.643511   0.312808   0.365473   0.148966   0.169884   \n",
       "896     0.595342    0.617808   0.180033   0.414860   0.145678   0.135584   \n",
       "897     0.512092    0.645780   0.451883   0.488395   0.233249   0.157940   \n",
       "898     0.805690    0.754852   0.210411   0.388109   0.211613   0.152250   \n",
       "899     0.305668    0.406003   0.212867   0.210097   0.055664   0.047573   \n",
       "\n",
       "     mfcc17_mean  \n",
       "0       0.425846  \n",
       "1       0.323413  \n",
       "2       0.475342  \n",
       "3       0.371888  \n",
       "4       0.158994  \n",
       "..           ...  \n",
       "895     0.406094  \n",
       "896     0.501825  \n",
       "897     0.535189  \n",
       "898     0.409805  \n",
       "899     0.369391  \n",
       "\n",
       "[900 rows x 32 columns]"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler = MinMaxScaler()\n",
    "scaler.fit(xc)\n",
    "xcm=pd.DataFrame(scaler.transform(xc),columns=xc.columns)\n",
    "scaler.fit(x)\n",
    "xm=pd.DataFrame(scaler.transform(x),columns=x.columns)\n",
    "xcm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "xm_train, xm_test, ym_train, ym_test = train_test_split(xcm, yc, test_size=0.20, random_state=42, stratify=yc)\n",
    "xmo_train, xmo_test, ymo_train, ymo_test = train_test_split(xm, yc, test_size=0.20, random_state=42, stratify=yc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Support Vector Machine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7235893867872384 {'C': 7.96, 'kernel': 'rbf'}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.705882</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.727273</td>\n",
       "      <td>0.650000</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.933333</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>0.789474</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.625000</td>\n",
       "      <td>0.766667</td>\n",
       "      <td>0.783652</td>\n",
       "      <td>0.783202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.705882</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.789474</td>\n",
       "      <td>0.766667</td>\n",
       "      <td>0.766202</td>\n",
       "      <td>0.766667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.705882</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.684211</td>\n",
       "      <td>0.642857</td>\n",
       "      <td>0.848485</td>\n",
       "      <td>0.787879</td>\n",
       "      <td>0.810811</td>\n",
       "      <td>0.789474</td>\n",
       "      <td>0.697674</td>\n",
       "      <td>0.766667</td>\n",
       "      <td>0.765616</td>\n",
       "      <td>0.765571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.766667</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.705882   0.888889   0.727273   0.650000   0.900000   0.933333   \n",
       "recall      0.705882   0.888889   0.888889   0.722222   0.500000   0.777778   \n",
       "f1-score    0.705882   0.888889   0.800000   0.684211   0.642857   0.848485   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.866667   0.789474   0.750000   0.625000  0.766667    0.783652   \n",
       "recall      0.722222   0.833333   0.833333   0.789474  0.766667    0.766202   \n",
       "f1-score    0.787879   0.810811   0.789474   0.697674  0.766667    0.765616   \n",
       "support    18.000000  18.000000  18.000000  19.000000  0.766667  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.783202  \n",
       "recall         0.766667  \n",
       "f1-score       0.765571  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with extracted features\n",
    "parameters = {'kernel':['rbf'], 'C':[7.96,7.97,7.98,7.99,8]}\n",
    "svc=svm.SVC()\n",
    "clf2b = GridSearchCV(svc, parameters, cv=5, scoring='f1_macro')\n",
    "clf2b.fit(xm_train,ym_train)\n",
    "print(clf2b.best_score_,clf2b.best_params_)\n",
    "ym_pred=clf2b.predict(xm_test)\n",
    "report=pd.DataFrame(classification_report(ym_test,ym_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7181934138996005 {'C': 7.92, 'kernel': 'rbf'}\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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</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>0.866667</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.625000</td>\n",
       "      <td>0.736842</td>\n",
       "      <td>0.769231</td>\n",
       "      <td>0.933333</td>\n",
       "      <td>0.882353</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.523810</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>0.775390</td>\n",
       "      <td>0.773485</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.764706</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.578947</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>0.762143</td>\n",
       "      <td>0.761111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1-score</th>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.714286</td>\n",
       "      <td>0.756757</td>\n",
       "      <td>0.645161</td>\n",
       "      <td>0.848485</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>0.789474</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.550000</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>0.764047</td>\n",
       "      <td>0.762589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>support</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.761111</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>180.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1          2          3          4          5  \\\n",
       "precision   0.866667   0.944444   0.625000   0.736842   0.769231   0.933333   \n",
       "recall      0.764706   0.944444   0.833333   0.777778   0.555556   0.777778   \n",
       "f1-score    0.812500   0.944444   0.714286   0.756757   0.645161   0.848485   \n",
       "support    17.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
       "\n",
       "                   6          7          8          9  accuracy   macro avg  \\\n",
       "precision   0.882353   0.750000   0.722222   0.523810  0.761111    0.775390   \n",
       "recall      0.833333   0.833333   0.722222   0.578947  0.761111    0.762143   \n",
       "f1-score    0.857143   0.789474   0.722222   0.550000  0.761111    0.764047   \n",
       "support    18.000000  18.000000  18.000000  19.000000  0.761111  180.000000   \n",
       "\n",
       "           weighted avg  \n",
       "precision      0.773485  \n",
       "recall         0.761111  \n",
       "f1-score       0.762589  \n",
       "support      180.000000  "
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with all features\n",
    "parameters = {'kernel':['rbf'], 'C':[7.92,7.93,7.94,7.95,7.96]}\n",
    "svc=svm.SVC()\n",
    "clf2bo = GridSearchCV(svc, parameters, cv=5, scoring='f1_macro')\n",
    "clf2bo.fit(xmo_train,ymo_train)\n",
    "print(clf2bo.best_score_,clf2bo.best_params_)\n",
    "ymo_pred=clf2bo.predict(xmo_test)\n",
    "report=pd.DataFrame(classification_report(ymo_test,ymo_pred,output_dict=True))\n",
    "report"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### FInal Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "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>chroma_stft_mean</th>\n",
       "      <th>length</th>\n",
       "      <th>perceptr_var</th>\n",
       "      <th>rolloff_mean</th>\n",
       "      <th>spectral_centroid_mean</th>\n",
       "      <th>spectral_bandwidth_mean</th>\n",
       "      <th>spectral_centroid_var</th>\n",
       "      <th>mfcc1_mean</th>\n",
       "      <th>rms_mean</th>\n",
       "      <th>rms_var</th>\n",
       "      <th>...</th>\n",
       "      <th>mfcc4_var</th>\n",
       "      <th>mfcc9_mean</th>\n",
       "      <th>harmony_var</th>\n",
       "      <th>mfcc11_mean</th>\n",
       "      <th>mfcc3_mean</th>\n",
       "      <th>mfcc7_var</th>\n",
       "      <th>mfcc4_var</th>\n",
       "      <th>mfcc5_var</th>\n",
       "      <th>mfcc1_var</th>\n",
       "      <th>mfcc17_mean</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.329507</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.012168</td>\n",
       "      <td>0.188901</td>\n",
       "      <td>0.177252</td>\n",
       "      <td>0.308429</td>\n",
       "      <td>0.246480</td>\n",
       "      <td>0.471850</td>\n",
       "      <td>0.183909</td>\n",
       "      <td>0.050489</td>\n",
       "      <td>...</td>\n",
       "      <td>0.457843</td>\n",
       "      <td>0.737134</td>\n",
       "      <td>0.051081</td>\n",
       "      <td>0.745266</td>\n",
       "      <td>0.667977</td>\n",
       "      <td>0.745105</td>\n",
       "      <td>0.457843</td>\n",
       "      <td>0.751786</td>\n",
       "      <td>0.530800</td>\n",
       "      <td>0.660392</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.622219</td>\n",
       "      <td>0.033658</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.095494</td>\n",
       "      <td>0.051192</td>\n",
       "      <td>0.256794</td>\n",
       "      <td>0.037277</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000623</td>\n",
       "      <td>...</td>\n",
       "      <td>0.034343</td>\n",
       "      <td>0.829382</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.760230</td>\n",
       "      <td>0.537419</td>\n",
       "      <td>0.165146</td>\n",
       "      <td>0.034343</td>\n",
       "      <td>0.066310</td>\n",
       "      <td>0.536294</td>\n",
       "      <td>0.780769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.673140</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.529442</td>\n",
       "      <td>0.611093</td>\n",
       "      <td>0.638172</td>\n",
       "      <td>0.607002</td>\n",
       "      <td>0.510081</td>\n",
       "      <td>0.837665</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.327139</td>\n",
       "      <td>...</td>\n",
       "      <td>0.324851</td>\n",
       "      <td>0.481280</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.631588</td>\n",
       "      <td>0.329212</td>\n",
       "      <td>0.496374</td>\n",
       "      <td>0.324851</td>\n",
       "      <td>0.874052</td>\n",
       "      <td>0.245020</td>\n",
       "      <td>0.338172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.370960</td>\n",
       "      <td>0.033658</td>\n",
       "      <td>0.006325</td>\n",
       "      <td>0.440487</td>\n",
       "      <td>0.330672</td>\n",
       "      <td>0.729691</td>\n",
       "      <td>0.394346</td>\n",
       "      <td>0.373225</td>\n",
       "      <td>0.144643</td>\n",
       "      <td>0.055743</td>\n",
       "      <td>...</td>\n",
       "      <td>0.260677</td>\n",
       "      <td>0.945142</td>\n",
       "      <td>0.043845</td>\n",
       "      <td>0.857139</td>\n",
       "      <td>0.911504</td>\n",
       "      <td>0.207466</td>\n",
       "      <td>0.260677</td>\n",
       "      <td>0.384891</td>\n",
       "      <td>0.506412</td>\n",
       "      <td>0.623321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.245510</td>\n",
       "      <td>0.033658</td>\n",
       "      <td>0.019243</td>\n",
       "      <td>0.454177</td>\n",
       "      <td>0.427353</td>\n",
       "      <td>0.435468</td>\n",
       "      <td>0.376193</td>\n",
       "      <td>0.565118</td>\n",
       "      <td>0.201190</td>\n",
       "      <td>0.083024</td>\n",
       "      <td>...</td>\n",
       "      <td>0.349495</td>\n",
       "      <td>0.508214</td>\n",
       "      <td>0.067338</td>\n",
       "      <td>0.220544</td>\n",
       "      <td>0.253516</td>\n",
       "      <td>0.966547</td>\n",
       "      <td>0.349495</td>\n",
       "      <td>0.942414</td>\n",
       "      <td>0.341319</td>\n",
       "      <td>0.081709</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",
       "      <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>95</th>\n",
       "      <td>0.374334</td>\n",
       "      <td>0.033658</td>\n",
       "      <td>0.092472</td>\n",
       "      <td>0.464699</td>\n",
       "      <td>0.486948</td>\n",
       "      <td>0.461878</td>\n",
       "      <td>0.570542</td>\n",
       "      <td>0.670103</td>\n",
       "      <td>0.274861</td>\n",
       "      <td>0.146595</td>\n",
       "      <td>...</td>\n",
       "      <td>0.248943</td>\n",
       "      <td>0.492267</td>\n",
       "      <td>0.073935</td>\n",
       "      <td>0.321861</td>\n",
       "      <td>0.288163</td>\n",
       "      <td>0.560631</td>\n",
       "      <td>0.248943</td>\n",
       "      <td>0.675645</td>\n",
       "      <td>0.240649</td>\n",
       "      <td>0.295719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>0.399430</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.210051</td>\n",
       "      <td>0.472771</td>\n",
       "      <td>0.460074</td>\n",
       "      <td>0.489646</td>\n",
       "      <td>0.111873</td>\n",
       "      <td>0.811391</td>\n",
       "      <td>0.424643</td>\n",
       "      <td>0.124319</td>\n",
       "      <td>...</td>\n",
       "      <td>0.225817</td>\n",
       "      <td>0.191832</td>\n",
       "      <td>0.122850</td>\n",
       "      <td>0.178145</td>\n",
       "      <td>0.112843</td>\n",
       "      <td>0.275701</td>\n",
       "      <td>0.225817</td>\n",
       "      <td>0.394567</td>\n",
       "      <td>0.197403</td>\n",
       "      <td>0.399098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>0.521726</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.476206</td>\n",
       "      <td>0.896350</td>\n",
       "      <td>0.901520</td>\n",
       "      <td>0.900221</td>\n",
       "      <td>0.361468</td>\n",
       "      <td>0.934745</td>\n",
       "      <td>0.741844</td>\n",
       "      <td>0.268320</td>\n",
       "      <td>...</td>\n",
       "      <td>0.368701</td>\n",
       "      <td>0.803309</td>\n",
       "      <td>0.343735</td>\n",
       "      <td>0.614670</td>\n",
       "      <td>0.557940</td>\n",
       "      <td>0.385687</td>\n",
       "      <td>0.368701</td>\n",
       "      <td>0.209057</td>\n",
       "      <td>0.300937</td>\n",
       "      <td>0.561906</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>0.608563</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.996808</td>\n",
       "      <td>0.740728</td>\n",
       "      <td>0.882498</td>\n",
       "      <td>0.840391</td>\n",
       "      <td>0.808508</td>\n",
       "      <td>...</td>\n",
       "      <td>0.512627</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.462038</td>\n",
       "      <td>0.748099</td>\n",
       "      <td>0.590660</td>\n",
       "      <td>0.444569</td>\n",
       "      <td>0.512627</td>\n",
       "      <td>0.867801</td>\n",
       "      <td>0.341452</td>\n",
       "      <td>0.556777</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>0.608534</td>\n",
       "      <td>0.033658</td>\n",
       "      <td>0.018657</td>\n",
       "      <td>0.658993</td>\n",
       "      <td>0.672095</td>\n",
       "      <td>0.598070</td>\n",
       "      <td>0.245328</td>\n",
       "      <td>0.670519</td>\n",
       "      <td>0.152369</td>\n",
       "      <td>0.018769</td>\n",
       "      <td>...</td>\n",
       "      <td>0.028246</td>\n",
       "      <td>0.061251</td>\n",
       "      <td>0.021028</td>\n",
       "      <td>0.312639</td>\n",
       "      <td>0.346462</td>\n",
       "      <td>0.082714</td>\n",
       "      <td>0.028246</td>\n",
       "      <td>0.051163</td>\n",
       "      <td>0.079837</td>\n",
       "      <td>0.378776</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    chroma_stft_mean    length  perceptr_var  rolloff_mean  \\\n",
       "0           0.329507  0.000000      0.012168      0.188901   \n",
       "1           0.622219  0.033658      0.000000      0.095494   \n",
       "2           0.673140  0.000000      0.529442      0.611093   \n",
       "3           0.370960  0.033658      0.006325      0.440487   \n",
       "4           0.245510  0.033658      0.019243      0.454177   \n",
       "..               ...       ...           ...           ...   \n",
       "95          0.374334  0.033658      0.092472      0.464699   \n",
       "96          0.399430  0.000000      0.210051      0.472771   \n",
       "97          0.521726  0.000000      0.476206      0.896350   \n",
       "98          0.608563  0.000000      1.000000      1.000000   \n",
       "99          0.608534  0.033658      0.018657      0.658993   \n",
       "\n",
       "    spectral_centroid_mean  spectral_bandwidth_mean  spectral_centroid_var  \\\n",
       "0                 0.177252                 0.308429               0.246480   \n",
       "1                 0.051192                 0.256794               0.037277   \n",
       "2                 0.638172                 0.607002               0.510081   \n",
       "3                 0.330672                 0.729691               0.394346   \n",
       "4                 0.427353                 0.435468               0.376193   \n",
       "..                     ...                      ...                    ...   \n",
       "95                0.486948                 0.461878               0.570542   \n",
       "96                0.460074                 0.489646               0.111873   \n",
       "97                0.901520                 0.900221               0.361468   \n",
       "98                1.000000                 0.996808               0.740728   \n",
       "99                0.672095                 0.598070               0.245328   \n",
       "\n",
       "    mfcc1_mean  rms_mean   rms_var  ...  mfcc4_var  mfcc9_mean  harmony_var  \\\n",
       "0     0.471850  0.183909  0.050489  ...   0.457843    0.737134     0.051081   \n",
       "1     0.000000  0.000000  0.000623  ...   0.034343    0.829382     0.000000   \n",
       "2     0.837665  1.000000  0.327139  ...   0.324851    0.481280     1.000000   \n",
       "3     0.373225  0.144643  0.055743  ...   0.260677    0.945142     0.043845   \n",
       "4     0.565118  0.201190  0.083024  ...   0.349495    0.508214     0.067338   \n",
       "..         ...       ...       ...  ...        ...         ...          ...   \n",
       "95    0.670103  0.274861  0.146595  ...   0.248943    0.492267     0.073935   \n",
       "96    0.811391  0.424643  0.124319  ...   0.225817    0.191832     0.122850   \n",
       "97    0.934745  0.741844  0.268320  ...   0.368701    0.803309     0.343735   \n",
       "98    0.882498  0.840391  0.808508  ...   0.512627    1.000000     0.462038   \n",
       "99    0.670519  0.152369  0.018769  ...   0.028246    0.061251     0.021028   \n",
       "\n",
       "    mfcc11_mean  mfcc3_mean  mfcc7_var  mfcc4_var  mfcc5_var  mfcc1_var  \\\n",
       "0      0.745266    0.667977   0.745105   0.457843   0.751786   0.530800   \n",
       "1      0.760230    0.537419   0.165146   0.034343   0.066310   0.536294   \n",
       "2      0.631588    0.329212   0.496374   0.324851   0.874052   0.245020   \n",
       "3      0.857139    0.911504   0.207466   0.260677   0.384891   0.506412   \n",
       "4      0.220544    0.253516   0.966547   0.349495   0.942414   0.341319   \n",
       "..          ...         ...        ...        ...        ...        ...   \n",
       "95     0.321861    0.288163   0.560631   0.248943   0.675645   0.240649   \n",
       "96     0.178145    0.112843   0.275701   0.225817   0.394567   0.197403   \n",
       "97     0.614670    0.557940   0.385687   0.368701   0.209057   0.300937   \n",
       "98     0.748099    0.590660   0.444569   0.512627   0.867801   0.341452   \n",
       "99     0.312639    0.346462   0.082714   0.028246   0.051163   0.079837   \n",
       "\n",
       "    mfcc17_mean  \n",
       "0      0.660392  \n",
       "1      0.780769  \n",
       "2      0.338172  \n",
       "3      0.623321  \n",
       "4      0.081709  \n",
       "..          ...  \n",
       "95     0.295719  \n",
       "96     0.399098  \n",
       "97     0.561906  \n",
       "98     0.556777  \n",
       "99     0.378776  \n",
       "\n",
       "[100 rows x 32 columns]"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#importing the test data\n",
    "tst=pd.read_csv('music30s_test.csv')\n",
    "tstc=tst[['chroma_stft_mean','length','perceptr_var','rolloff_mean','spectral_centroid_mean','spectral_bandwidth_mean','spectral_centroid_var','mfcc1_mean','rms_mean','rms_var','mfcc2_mean','mfcc4_mean','rolloff_var','mfcc12_mean','mfcc6_var','chroma_stft_var','mfcc20_var','mfcc8_mean','mfcc7_mean','perceptr_mean','mfcc6_mean','zero_crossing_rate_mean','mfcc4_var','mfcc9_mean','harmony_var','mfcc11_mean','mfcc3_mean','mfcc7_var','mfcc4_var','mfcc5_var','mfcc1_var','mfcc17_mean']].copy()\n",
    "scaler = MinMaxScaler()\n",
    "scaler.fit(tstc)\n",
    "tstcm=pd.DataFrame(scaler.transform(tstc),columns=tstc.columns)\n",
    "tstcm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python39\\lib\\site-packages\\sklearn\\utils\\validation.py:63: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  return f(*args, **kwargs)\n"
     ]
    }
   ],
   "source": [
    "# using svm on trimmed and scaled test data\n",
    "lab=clf2b.predict(tstcm)\n",
    "f_labels=pd.DataFrame({'label':lab})\n",
    "f_labels=le.inverse_transform(f_labels)\n",
    "f_labels=pd.DataFrame({'label':f_labels})\n",
    "final=f_labels.to_numpy()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "#generating the final text labels\n",
    "np.savetxt(r'Marut_Priyadarshi_Label.txt',final,fmt='%s')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "8da01e5a71448ea74f54d88afa8911010d1d12e23bc7e103d40d5def4a09152c"
  },
  "kernelspec": {
   "display_name": "Python 3.9.6 64-bit",
   "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": "3.9.6"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}