NB_SGD_KNN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "id": "bSqk3eqNcEJI"
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "id": "6mUu-3ruclOI"
   },
   "outputs": [],
   "source": [
    "df=pd.read_csv('datasets/GTZAN/features_3_sec.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "id": "KduTbN1dhyDS"
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import LabelEncoder\n",
    "le=LabelEncoder()\n",
    "df['label']=le.fit_transform(df['label']) #encoding the label column ie the genre"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "zMjuhmkEdFGO",
    "outputId": "ab137974-7af9-42bd-f526-84264a2c50fc"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['filename', 'length', 'chroma_stft_mean', 'chroma_stft_var', 'rms_mean', 'rms_var', 'spectral_centroid_mean', 'spectral_centroid_var', 'spectral_bandwidth_mean', 'spectral_bandwidth_var', 'rolloff_mean', 'rolloff_var', 'zero_crossing_rate_mean', 'zero_crossing_rate_var', 'harmony_mean', 'harmony_var', 'perceptr_mean', 'perceptr_var', 'tempo', 'mfcc1_mean', 'mfcc1_var', 'mfcc2_mean', 'mfcc2_var', 'mfcc3_mean', 'mfcc3_var', 'mfcc4_mean', 'mfcc4_var', 'mfcc5_mean', 'mfcc5_var', 'mfcc6_mean', 'mfcc6_var', 'mfcc7_mean', 'mfcc7_var', 'mfcc8_mean', 'mfcc8_var', 'mfcc9_mean', 'mfcc9_var', 'mfcc10_mean', 'mfcc10_var', 'mfcc11_mean', 'mfcc11_var', 'mfcc12_mean', 'mfcc12_var', 'mfcc13_mean', 'mfcc13_var', 'mfcc14_mean', 'mfcc14_var', 'mfcc15_mean', 'mfcc15_var', 'mfcc16_mean', 'mfcc16_var', 'mfcc17_mean', 'mfcc17_var', 'mfcc18_mean', 'mfcc18_var', 'mfcc19_mean', 'mfcc19_var', 'mfcc20_mean', 'mfcc20_var', 'label']\n",
      "['length', 'chroma_stft_mean', 'chroma_stft_var', 'rms_mean', 'rms_var', 'spectral_centroid_mean', 'spectral_centroid_var', 'spectral_bandwidth_mean', 'spectral_bandwidth_var', 'rolloff_mean', 'rolloff_var', 'zero_crossing_rate_mean', 'zero_crossing_rate_var', 'harmony_mean', 'harmony_var', 'perceptr_mean', 'perceptr_var', 'tempo', 'mfcc1_mean', 'mfcc1_var', 'mfcc2_mean', 'mfcc2_var', 'mfcc3_mean', 'mfcc3_var', 'mfcc4_mean', 'mfcc4_var', 'mfcc5_mean', 'mfcc5_var', 'mfcc6_mean', 'mfcc6_var', 'mfcc7_mean', 'mfcc7_var', 'mfcc8_mean', 'mfcc8_var', 'mfcc9_mean', 'mfcc9_var', 'mfcc10_mean', 'mfcc10_var', 'mfcc11_mean', 'mfcc11_var', 'mfcc12_mean', 'mfcc12_var', 'mfcc13_mean', 'mfcc13_var', 'mfcc14_mean', 'mfcc14_var', 'mfcc15_mean', 'mfcc15_var', 'mfcc16_mean', 'mfcc16_var', 'mfcc17_mean', 'mfcc17_var', 'mfcc18_mean', 'mfcc18_var', 'mfcc19_mean', 'mfcc19_var', 'mfcc20_mean', 'mfcc20_var']\n",
      "58\n"
     ]
    }
   ],
   "source": [
    "cols=list(df.columns)\n",
    "print(cols)\n",
    "train_cols=cols[1:-1]\n",
    "test_cols=cols[-1]\n",
    "print(train_cols)\n",
    "print(len(train_cols))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 223
    },
    "id": "WDwAPJYpwGEP",
    "outputId": "07026cd5-74ea-4702-8a9e-50ed686f0c90"
   },
   "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>filename</th>\n",
       "      <th>length</th>\n",
       "      <th>chroma_stft_mean</th>\n",
       "      <th>chroma_stft_var</th>\n",
       "      <th>rms_mean</th>\n",
       "      <th>rms_var</th>\n",
       "      <th>spectral_centroid_mean</th>\n",
       "      <th>spectral_centroid_var</th>\n",
       "      <th>spectral_bandwidth_mean</th>\n",
       "      <th>spectral_bandwidth_var</th>\n",
       "      <th>...</th>\n",
       "      <th>mfcc16_var</th>\n",
       "      <th>mfcc17_mean</th>\n",
       "      <th>mfcc17_var</th>\n",
       "      <th>mfcc18_mean</th>\n",
       "      <th>mfcc18_var</th>\n",
       "      <th>mfcc19_mean</th>\n",
       "      <th>mfcc19_var</th>\n",
       "      <th>mfcc20_mean</th>\n",
       "      <th>mfcc20_var</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>blues.00000.0.wav</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.487808</td>\n",
       "      <td>0.640520</td>\n",
       "      <td>-0.006624</td>\n",
       "      <td>0.235568</td>\n",
       "      <td>-0.566828</td>\n",
       "      <td>-0.572791</td>\n",
       "      <td>-0.493983</td>\n",
       "      <td>-0.009229</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.299108</td>\n",
       "      <td>0.168647</td>\n",
       "      <td>-0.425137</td>\n",
       "      <td>-0.003423</td>\n",
       "      <td>-0.376938</td>\n",
       "      <td>-0.499464</td>\n",
       "      <td>-0.513562</td>\n",
       "      <td>0.128414</td>\n",
       "      <td>-0.291781</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>blues.00000.1.wav</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.403142</td>\n",
       "      <td>0.131835</td>\n",
       "      <td>-0.264944</td>\n",
       "      <td>-0.342134</td>\n",
       "      <td>-0.508798</td>\n",
       "      <td>-0.749862</td>\n",
       "      <td>-0.425382</td>\n",
       "      <td>-0.519010</td>\n",
       "      <td>...</td>\n",
       "      <td>0.428544</td>\n",
       "      <td>-0.327031</td>\n",
       "      <td>-0.310040</td>\n",
       "      <td>-0.112125</td>\n",
       "      <td>-0.032083</td>\n",
       "      <td>-0.066593</td>\n",
       "      <td>1.011384</td>\n",
       "      <td>1.275780</td>\n",
       "      <td>0.056425</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>blues.00000.2.wav</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.361694</td>\n",
       "      <td>0.764491</td>\n",
       "      <td>0.016695</td>\n",
       "      <td>0.542195</td>\n",
       "      <td>-0.546245</td>\n",
       "      <td>-0.701852</td>\n",
       "      <td>-0.288365</td>\n",
       "      <td>-0.425734</td>\n",
       "      <td>...</td>\n",
       "      <td>0.503695</td>\n",
       "      <td>0.428053</td>\n",
       "      <td>-0.648762</td>\n",
       "      <td>0.316311</td>\n",
       "      <td>-0.177372</td>\n",
       "      <td>0.109337</td>\n",
       "      <td>-0.046244</td>\n",
       "      <td>0.653907</td>\n",
       "      <td>-0.521458</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>blues.00000.3.wav</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.175714</td>\n",
       "      <td>0.205477</td>\n",
       "      <td>0.024885</td>\n",
       "      <td>-0.063820</td>\n",
       "      <td>-0.723482</td>\n",
       "      <td>-0.700599</td>\n",
       "      <td>-0.517344</td>\n",
       "      <td>-0.348881</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.065309</td>\n",
       "      <td>0.062981</td>\n",
       "      <td>-0.649076</td>\n",
       "      <td>0.092384</td>\n",
       "      <td>-0.464121</td>\n",
       "      <td>-0.211882</td>\n",
       "      <td>-0.099501</td>\n",
       "      <td>0.865880</td>\n",
       "      <td>-0.544744</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>blues.00000.4.wav</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.485895</td>\n",
       "      <td>0.337521</td>\n",
       "      <td>0.181345</td>\n",
       "      <td>-0.272072</td>\n",
       "      <td>-0.756246</td>\n",
       "      <td>-0.774827</td>\n",
       "      <td>-0.538557</td>\n",
       "      <td>-0.572962</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.570609</td>\n",
       "      <td>0.856651</td>\n",
       "      <td>-0.167089</td>\n",
       "      <td>0.183265</td>\n",
       "      <td>-0.029476</td>\n",
       "      <td>-0.175130</td>\n",
       "      <td>-0.678995</td>\n",
       "      <td>0.276899</td>\n",
       "      <td>-0.606692</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 60 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            filename  length  chroma_stft_mean  chroma_stft_var  rms_mean  \\\n",
       "0  blues.00000.0.wav     0.0         -0.487808         0.640520 -0.006624   \n",
       "1  blues.00000.1.wav     0.0         -0.403142         0.131835 -0.264944   \n",
       "2  blues.00000.2.wav     0.0         -0.361694         0.764491  0.016695   \n",
       "3  blues.00000.3.wav     0.0         -0.175714         0.205477  0.024885   \n",
       "4  blues.00000.4.wav     0.0         -0.485895         0.337521  0.181345   \n",
       "\n",
       "    rms_var  spectral_centroid_mean  spectral_centroid_var  \\\n",
       "0  0.235568               -0.566828              -0.572791   \n",
       "1 -0.342134               -0.508798              -0.749862   \n",
       "2  0.542195               -0.546245              -0.701852   \n",
       "3 -0.063820               -0.723482              -0.700599   \n",
       "4 -0.272072               -0.756246              -0.774827   \n",
       "\n",
       "   spectral_bandwidth_mean  spectral_bandwidth_var  ...  mfcc16_var  \\\n",
       "0                -0.493983               -0.009229  ...   -0.299108   \n",
       "1                -0.425382               -0.519010  ...    0.428544   \n",
       "2                -0.288365               -0.425734  ...    0.503695   \n",
       "3                -0.517344               -0.348881  ...   -0.065309   \n",
       "4                -0.538557               -0.572962  ...   -0.570609   \n",
       "\n",
       "   mfcc17_mean  mfcc17_var  mfcc18_mean  mfcc18_var  mfcc19_mean  mfcc19_var  \\\n",
       "0     0.168647   -0.425137    -0.003423   -0.376938    -0.499464   -0.513562   \n",
       "1    -0.327031   -0.310040    -0.112125   -0.032083    -0.066593    1.011384   \n",
       "2     0.428053   -0.648762     0.316311   -0.177372     0.109337   -0.046244   \n",
       "3     0.062981   -0.649076     0.092384   -0.464121    -0.211882   -0.099501   \n",
       "4     0.856651   -0.167089     0.183265   -0.029476    -0.175130   -0.678995   \n",
       "\n",
       "   mfcc20_mean  mfcc20_var  label  \n",
       "0     0.128414   -0.291781      0  \n",
       "1     1.275780    0.056425      0  \n",
       "2     0.653907   -0.521458      0  \n",
       "3     0.865880   -0.544744      0  \n",
       "4     0.276899   -0.606692      0  \n",
       "\n",
       "[5 rows x 60 columns]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "scaler=StandardScaler()\n",
    "scaler.fit(df[train_cols]) #Normalizing the given data\n",
    "df[train_cols]=scaler.transform(df[train_cols])\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "R-VPLz3lkwBt",
    "outputId": "313caec3-7773-4b5f-c532-3abfe3f3ed6d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[False  True  True  True  True  True  True  True False  True  True  True\n",
      "  True False False False  True False  True False  True False  True False\n",
      "  True  True  True  True  True  True  True  True  True False  True False\n",
      "  True False False False  True False  True False False False  True False\n",
      " False False  True False False False False False False False]\n",
      "['chroma_stft_mean', 'chroma_stft_var', 'rms_mean', 'rms_var', 'spectral_centroid_mean', 'spectral_centroid_var', 'spectral_bandwidth_mean', 'rolloff_mean', 'rolloff_var', 'zero_crossing_rate_mean', 'zero_crossing_rate_var', 'perceptr_var', 'mfcc1_mean', 'mfcc2_mean', 'mfcc3_mean', 'mfcc4_mean', 'mfcc4_var', 'mfcc5_mean', 'mfcc5_var', 'mfcc6_mean', 'mfcc6_var', 'mfcc7_mean', 'mfcc7_var', 'mfcc8_mean', 'mfcc9_mean', 'mfcc10_mean', 'mfcc12_mean', 'mfcc13_mean', 'mfcc15_mean', 'mfcc17_mean']\n",
      "(7992, 30)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/feature_selection/_univariate_selection.py:112: UserWarning: Features [0] are constant.\n",
      "  warnings.warn(\"Features %s are constant.\" % constant_features_idx, UserWarning)\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/feature_selection/_univariate_selection.py:113: RuntimeWarning: invalid value encountered in true_divide\n",
      "  f = msb / msw\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test=train_test_split(df[train_cols],df[test_cols],test_size=0.2)\n",
    "from sklearn.feature_selection import f_classif, SelectKBest\n",
    "sel_f = SelectKBest(f_classif, k=30)    # select 30 features\n",
    "X_train_chi2 = sel_f.fit_transform(X_train, y_train) #selecting best 30 features\n",
    "print(sel_f.get_support())\n",
    "lsb=sel_f.get_support()\n",
    "ls=[]\n",
    "for i in range(len(train_cols)):\n",
    "  if lsb[i]==True:\n",
    "    ls.append(train_cols[i])\n",
    "print(ls)\n",
    "train_cols=ls\n",
    "X_train, X_test, y_train, y_test=train_test_split(df[train_cols],df[test_cols],test_size=0.2)\n",
    "print(X_train.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "id": "inu0YsT4-IKG"
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix, precision_score, accuracy_score, recall_score, classification_report\n",
    "def model_assess(model, title = \"Default\"):\n",
    "    y_pred = model.predict(X_test)\n",
    "    print(title, ':')\n",
    "    # # predict classes on testing data\n",
    "    cm = confusion_matrix(y_test, y_pred)\n",
    "\n",
    "    # # plotting conf matrix using seaborn\n",
    "    plt.figure(figsize = (10,7))\n",
    "    mask = np.triu(np.ones_like(cm, dtype=np.bool))\n",
    "    sns.heatmap(cm, mask=mask, vmax=30, center=0,annot=True)\n",
    "    # plt.savefig(\"\")\n",
    "\n",
    "    # Predict accuracy,precision_score,recall_score\n",
    "    print(f\"Accuracy score: {accuracy_score(y_test, y_pred)}\")\n",
    "    print(f\"Overall Precision: {precision_score(y_test, y_pred,average='macro')}\")\n",
    "    print(f\"Overall Recall:{recall_score(y_test, y_pred,average='macro')}\")\n",
    "    # Print classification Report\n",
    "    print(classification_report(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "6TyBQzGZ_7kt"
   },
   "source": [
    "# K Nearest Neighbors Classifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "vad5SaqbBNKk"
   },
   "source": [
    "## Basic sklearn KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "7c78sskwAKZk",
    "outputId": "181d5604-8e08-4b99-c65b-6c9a91f83821"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Default :\n",
      "Accuracy score: 0.8708708708708709\n",
      "Overall Precision: 0.8733693484899305\n",
      "Overall Recall:0.8711782026612752\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.88      0.89      0.89       191\n",
      "           1       0.90      0.96      0.93       214\n",
      "           2       0.77      0.84      0.80       193\n",
      "           3       0.82      0.87      0.84       202\n",
      "           4       0.92      0.87      0.89       201\n",
      "           5       0.87      0.86      0.86       202\n",
      "           6       0.96      0.92      0.94       187\n",
      "           7       0.88      0.84      0.86       177\n",
      "           8       0.84      0.90      0.87       211\n",
      "           9       0.89      0.76      0.82       220\n",
      "\n",
      "    accuracy                           0.87      1998\n",
      "   macro avg       0.87      0.87      0.87      1998\n",
      "weighted avg       0.87      0.87      0.87      1998\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "simpleknn=KNeighborsClassifier(n_neighbors=5) #only 5 neighbors\n",
    "simpleknn.fit(X_train,y_train)\n",
    "model_assess(simpleknn)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "5xHkZDAcBSgP"
   },
   "source": [
    "## Using Grid Search CV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "WWvjw6BHxRWw",
    "outputId": "972f13aa-2c7a-4691-e9c6-cadd33725f7d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 76 candidates, totalling 380 fits\n",
      "{'metric': 'manhattan', 'n_neighbors': 1, 'weights': 'uniform'}\n"
     ]
    }
   ],
   "source": [
    "n=len(set(y_train))\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "params_KNN = {'weights': ['uniform','distance'],\n",
    "              'metric': ['euclidean','manhattan'],\n",
    "              'n_neighbors':list(range(1,20))\n",
    "              }                                     #setting param grid\n",
    "modelk=KNeighborsClassifier()\n",
    "gs_KNN=GridSearchCV(estimator=modelk,\n",
    "                    param_grid=params_KNN,\n",
    "                    verbose=1, \n",
    "                    scoring='accuracy')\n",
    "gs_KNN.fit(X_train,y_train)\n",
    "print(gs_KNN.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "U7LdCiNC-i7c",
    "outputId": "25241507-8f9b-40ad-8235-94f3132a75d1"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNN GridsearchCV :\n",
      "Accuracy score: 0.9059059059059059\n",
      "Overall Precision: 0.9073617032054686\n",
      "Overall Recall:0.905944266718195\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.93      0.90      0.91       191\n",
      "           1       0.94      0.96      0.95       214\n",
      "           2       0.82      0.82      0.82       193\n",
      "           3       0.87      0.91      0.89       202\n",
      "           4       0.95      0.94      0.95       201\n",
      "           5       0.91      0.90      0.91       202\n",
      "           6       0.96      0.97      0.97       187\n",
      "           7       0.95      0.89      0.92       177\n",
      "           8       0.89      0.94      0.91       211\n",
      "           9       0.85      0.83      0.84       220\n",
      "\n",
      "    accuracy                           0.91      1998\n",
      "   macro avg       0.91      0.91      0.91      1998\n",
      "weighted avg       0.91      0.91      0.91      1998\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model_assess(gs_KNN.best_estimator_,\"KNN GridsearchCV\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Fut42RgbBqnI"
   },
   "source": [
    "## Drawing the ROC curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 295
    },
    "id": "uRh0orAZ0hA4",
    "outputId": "f1ef871d-554e-4cad-d6f5-1ddad2fc9148"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_curve\n",
    "def ROC(model):\n",
    "    pred_prob = model.predict_proba(X_test)\n",
    "    classes=['blues','classical','country','disco','hiphop','jazz','metal','pop','reggae','rock']\n",
    "    # Get ROC metrics for each class\n",
    "    fpr = {}\n",
    "    tpr = {}\n",
    "    thresh ={}\n",
    "    for i in range(len(classes)):\n",
    "        fpr[i], tpr[i], thresh[i] = roc_curve(y_test, pred_prob[:,i], pos_label=i)\n",
    "\n",
    "    # Plot the ROC chart\n",
    "    for i in range(len(classes)):\n",
    "        plt.plot(fpr[i], tpr[i], linestyle='--', label=classes[i] + ' vs Rest')\n",
    "\n",
    "    plt.title('Multiclass ROC curve')\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.ylabel('True Positive rate')\n",
    "    plt.legend(loc='best')\n",
    "    plt.show()\n",
    "ROC(gs_KNN.best_estimator_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "rZmA9a4BCN4b"
   },
   "source": [
    "# Gaussian Naive Bayes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "XrCX877ACUbD"
   },
   "source": [
    "## Simple and One vs Rest Gaussian NB"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "vTWci2wzmFaq",
    "outputId": "a54d2b24-06a6-42ef-9e55-6cd570ff594b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Simple Gaussian NB :\n",
      "Accuracy score: 0.46896896896896895\n",
      "Overall Precision: 0.4606482379884576\n",
      "Overall Recall:0.4755693757933862\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.39      0.46      0.42       191\n",
      "           1       0.68      0.88      0.77       214\n",
      "           2       0.34      0.29      0.31       193\n",
      "           3       0.39      0.27      0.32       202\n",
      "           4       0.65      0.30      0.41       201\n",
      "           5       0.49      0.39      0.43       202\n",
      "           6       0.45      0.87      0.59       187\n",
      "           7       0.45      0.76      0.57       177\n",
      "           8       0.52      0.41      0.46       211\n",
      "           9       0.26      0.12      0.17       220\n",
      "\n",
      "    accuracy                           0.47      1998\n",
      "   macro avg       0.46      0.48      0.44      1998\n",
      "weighted avg       0.46      0.47      0.44      1998\n",
      "\n",
      "One Vs Rest :\n",
      "Accuracy score: 0.46646646646646645\n",
      "Overall Precision: 0.45681941758051003\n",
      "Overall Recall:0.47323250823669855\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.39      0.46      0.42       191\n",
      "           1       0.66      0.91      0.77       214\n",
      "           2       0.33      0.26      0.29       193\n",
      "           3       0.38      0.25      0.30       202\n",
      "           4       0.64      0.27      0.38       201\n",
      "           5       0.49      0.38      0.43       202\n",
      "           6       0.44      0.89      0.59       187\n",
      "           7       0.44      0.78      0.56       177\n",
      "           8       0.52      0.42      0.46       211\n",
      "           9       0.27      0.11      0.16       220\n",
      "\n",
      "    accuracy                           0.47      1998\n",
      "   macro avg       0.46      0.47      0.44      1998\n",
      "weighted avg       0.46      0.47      0.43      1998\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiAAAAGfCAYAAABm/WkhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3deXhU5dnH8e+dhC1hR5awKLhVUAElKqIFFPGlKi/uFcGtVNRara1L1VZQ0KKiqH1roRRUqEUWRVaLAooUd0BWUdkFSdgCQsKWTJ73j4wRNMlMcHKeHPx9rmsuZs7MnPMjVya589zPeY455xAREREJUpLvACIiIvLTowJEREREAqcCRERERAKnAkREREQCpwJEREREAqcCRERERAKnAkRERETiYmZVzexjM1tsZsvN7JHo9hZm9pGZrTSzcWZWOda+VICIiIhIvPYD5zvn2gBtgW5m1h54AnjGOXcCsAPoE2tHKkBEREQkLq5QTvRhpejNAecDr0a3jwIujbWvlHJJeCgttSoiIj81FujRlr+WsN+1dsqVtwB9D9o03Dk3vOh5s2RgAXA88DywGtjpnMuPvmQj0CTWcYIoQERERCQkosXG8FKejwBtzaw28DrQsriXxTqOWjAiIiJSZs65ncAcoD1Q28y+HdRoCmyK9X4VICIiIhIXM6sfHfnAzKoBFwArgHeAK6MvuwGYHGtfasGIiIiEnItEEravGJNX0oFR0XkgScB459w0M/sMGGtmjwKfAiNjHUcFiIiIiMTFObcEOK2Y7WuAM8uyL7VgREREJHAaAREREQm7SH7s11QwGgERERGRwKkAERERkcCpBSMiIhJyriBxLZiglnDVCIiIiIgETgWIiIiIBE4tGBERkbBL4EJkQdEIiIiIiAROBYiIiIgETi0YERGRkHMhXIgsZgFiZicBPYAmgKPwErtTnHMryjmbiIiIHKFKbcGY2R+BsRSeFvwx8En0/itmdn/5xxMREZEjUawRkD7Ayc65vIM3mtkQYDnweHFvMrO+QF+Af/zjH/Tt2zcBUUVERKRYR2ALpgBoDKz/3vb06HPFcs4NB4Z/+/Cw04mIiMgRKVYBchcw28xWAhui244Gjgd+W57BRERE5MhVagHinJthZicCZ1I4CdWAjcAnzrnwrXoiIiJyBErktWCCEvMsGOdcAfBhAFlERETkJ0ILkYmIiEjgtBCZiIhI2OlaMCIiIiKxqQARERGRwKkFIyIiEnJhvBaMRkBEREQkcCpAREREJHBqwYiIiISdWjAiIiIisWkEREREJORcgdYBEREREYlJBYiIiIgETi0YERGRkNM6ICIiIiJxKPcRkFbHtCjvQyTcZ+vX+o4gIiJyRFMLRkREJOzUghERERGJTQWIiIiIBE4tGBERkZDTQmQiIiIicVABIiIiIoFTC0ZERCTsdBaMiIiISGwqQERERCRwasGIiIiEnK4FIyIiIhIHFSAiIiISOLVgREREwk4tGBEREZHYQjcC0ig9nUHPPM1R9evjCgoYP+YVXn7xJd+xREREpAxCV4DkR/J58tHHWLFsOalpabw6bSofzJvH6pWrfEcTERHxQteCCcC2LVtZsWw5AHtyc1mzahUNGjbynEpERETKInQFyMEaN21Cy5NbsWTRIt9RREREpAwOuwAxs5tKea6vmc03s/k7cnYf7iFKlZqaynPDhjJowEByc3LK5RgiIiKhEMlP3C0gP2YE5JGSnnDODXfOZTjnMupUr/EjDlG8lJQUnh02lGmTJjNrxpsJ37+IiIiUr1InoZrZkpKeAhomPk58Bj75BGtWrWLUiJG+IoiIiMiPEOssmIbA/wA7vrfdgPfLJVEMp2dk0OOKy/lixedMfGM6AM8OHszcd+b4iCMiIuKdi4TvLJhYBcg0oLpz7gezPM1sTrkkimHh/Pm0OqaFj0OLiIj8pJlZM2A00AgoAIY7554zs4eBm4Gt0Zc+6Jx7o7R9lVqAOOf6lPLctWUJLSIiIqGXD9ztnFtoZjWABWY2M/rcM865p+LdUegWIhMREZFDuYDOXnHOZQKZ0fu7zWwF0ORw9hXqdUBEREQksQ5eSiN661vC65oDpwEfRTf91syWmNkLZlYn1nFUgIiIiEiRg5fSiN6Gf/81ZlYdeA24yzm3CxgKHAe0pXCE5OlYx1ELRkREJOwKgltAzMwqUVh8/Ns5NxHAObf5oOf/SeFJLKXSCIiIiIjExcwMGAmscM4NOWh7+kEvuwxYFmtfGgERERGReJ0DXAcsNbNvl+h4EOhpZm0BB6wDbom1IxUgIiIiIRfUQmTOuXkULkb6faWu+VEctWBEREQkcCpAREREJHBqwYiIiIRdCK8FoxEQERERCZwKEBEREQmcWjAiIiIhF9S1YBKp3AuQz96I+8J4FcID1/wfD5za2XeMMhu0dI7vCCIiInFTC0ZEREQCpxaMiIhI2OksGBEREZHYVICIiIhI4NSCERERCbmgrgWTSBoBERERkcCpABEREZHAqQUjIiIScq5ALRgRERGRmFSAiIiISODUghEREQk7nQUjIiIiEpsKEBEREQmcWjAiIiIhF8aFyEJTgJx/y5OkVatCUlISyclJTBx8O/95fyl/Gzeb1Ru3MuGJ2zj1+Ka+Yxa5YsB9nNTxbHKyd/Lc5TcBUK1mDXo+1Z86jRuxY1MWY+55mH27cjwnFRERCV6oWjCjBvyayUPuYOLg2wE48eiG/N99vTijVXO/wYqxYPIMXrztvkO2depzLas/WsjTl/Rm9UcL6dznWk/pRERE/ApVAfJ9xzVtwLFN6vuOUax1C5aw55vdh2xrdd45LJw8A4CFk2fQ6rxzfUQTEZEjjIsUJOwWlJgtGDM7CWgCfOScyzloezfn3IzyDPe9IPR55EXM4JcXnskvLzwzsEMnSvV6ddm9LRuA3duyqV6vjudEIiIifpQ6AmJmdwKTgTuAZWbW46Cn/1LK+/qa2Xwzmz98wsyEBH3lL7fw+tO/5Z9/vpF//+dDPlm+NiH7FRERkeDFGgG5GWjnnMsxs+bAq2bW3Dn3HGAlvck5NxwYDsDy11wigjasWxOAerWr0/WsVixZuZEzTm6RiF0HJmd7NjWOKhwFqXFUXXK27/AdSUREjgQBtk4SJdYckORv2y7OuXVAZ+AXZjaEUgqQRNuz7wA5e/cX3X9v8SpOOLphUIdPmBVz3uf0Ht0AOL1HNz575z3PiURERPyINQKSZWZtnXOLAKIjIZcALwCnlnu6qO07c7j9iZcBiBQUcMnP29Dx9BOZ+eFyBo6YSvauXG55bBQtWzRmZL+bgopVqmueeIgWZ7QlrXYt7p81gVnPv8i7I8fQ86n+ZFx2ETszNzPm7od9xxQREfHCnCu5Q2JmTYF851xWMc+d45yL/Sd8glowQXngmv/zHeGwDFo6x3cEERH5TmBdAoDtD5ybsN+19QbNCyR7qSMgzrmNpTyn/oGIiIgcllCvAyIiIiLhFJql2EVERKR4LhKq2Q6ARkBERETEAxUgIiIiEji1YEREREIuyGu4JIpGQERERCRwKkBEREQkcGrBiIiIhJxaMCIiIiJxUAEiIiIigVMLRkREJORcgRYiExEREYlJBYiIiIgErtxbMGG7vP3kXet9Ryiz2smV6XDsz3zHKLP313zhO4KIyBFB14IRERERiYMKEBEREQmczoIREREJORfxnaDsNAIiIiIigVMBIiIiIoFTC0ZERCTkdBaMiIiISBxUgIiIiEjg1IIREREJuYIC3wnKTiMgIiIiEhcza2Zm75jZCjNbbma/i26va2YzzWxl9N86sfalAkRERETilQ/c7ZxrCbQHbjezVsD9wGzn3AnA7OjjUqkFIyIiEnJBLUTmnMsEMqP3d5vZCqAJ0APoHH3ZKGAO8MfS9qUREBERESliZn3NbP5Bt74lvK45cBrwEdAwWpx8W6Q0iHUcjYCIiIhIEefccGB4aa8xs+rAa8BdzrldZlbm44SiALliwH2c1PFscrJ38tzlNwFQrWYNej7VnzqNG7FjUxZj7nmYfbtyPCctXqP0dAY98zRH1a+PKyhg/JhXePnFl3zHiikpKYkXJr/G1s2buffXt/qOIyIiJQjyWjBmVonC4uPfzrmJ0c2bzSzdOZdpZunAllj7CUULZsHkGbx4232HbOvU51pWf7SQpy/pzeqPFtK5z7We0sWWH8nnyUcfo3uXrlxz6eVce/31HHfC8b5jxXT1TdezbvVq3zFERKSCsMKhjpHACufckIOemgLcEL1/AzA51r5CUYCsW7CEPd/sPmRbq/POYeHkGQAsnDyDVued6yNaXLZt2cqKZcsB2JOby5pVq2jQsJHnVKWr36ghHc7rzNRxr/qOIiIiFcc5wHXA+Wa2KHq7CHgc6GpmK4Gu0celitmCMbMzAeec+yR6qk034HPn3Bs/6r/wI1WvV5fd27IB2L0tm+r1Yp5yXCE0btqElie3YsmiRb6jlOquhx7k+ccHk5qW5juKiIjEENRCZM65eUBJEz66lGVfpY6AmFl/4K/AUDMbBPwNqA7cb2Z/KuV9RTNoF2VvKkueI1pqairPDRvKoAEDyc2pmPNVADqc35kd27P5IjpqIyIikmixRkCuBNoCVYAsoGl0tutgCk+7eay4Nx08g/aBUzuXyyX6crZnU+OowlGQGkfVJWf7jvI4TMKkpKTw7LChTJs0mVkz3vQdp1St253OuV3O5+zOHalcpQpp1avTf8hgHvnDvb6jiYjIESLWHJB851zEObcHWO2c2wXgnNsLeF15fsWc9zm9RzcATu/Rjc/eec9nnJgGPvkEa1atYtSIkb6jxDRs8BAuPacTV3TsQr87/8CCDz5U8SEiUoG5SOJuQYlVgBwws9To/XbfbjSzWgRYgFzzxEPc9vLz1G/ejPtnTSDjsot4d+QYjm/fjrunvczx7dvx7sgxQcUps9MzMuhxxeWc1aEDE9+YzsQ3ptPxvM6+Y4mIiHhjzpXcITGzKs65/cVsPwpId84tjXWA8mrBlJfJu9b7jlBmtZMr+45wWN5f84XvCCIi5aXsK3P9CGuvyUjY79oWY+cHkr3UOSDFFR/R7duAbeWSSERERMqkoCDQeichQrEOiIiIiBxZVICIiIhI4EJxLRgREREpWVALkSWSRkBEREQkcBoBERERCbkg1+9IFI2AiIiISOBUgIiIiEjg1IIREREJOa0DIiIiIhIHFSAiIiISOLVgREREQq5AZ8GIiIiIxKYCRERERAJX7i2YXJdf3odIqAOEbz3bLZF9viOU2eN1T+DV0y7wHaNMrvx0lu8IIiLF0lkwIiIiInFQASIiIiKB01kwIiIiIefUghERERGJTQWIiIiIBE4tGBERkZArCN8JnBoBERERkeCpABEREZHAqQUjIiISclqITERERCQOKkBEREQkcGrBiIiIhJxaMCIiIiJxUAEiIiIigQtlAdKp91Xc//poHpj0Lzr3vsp3nJgGDR7MRwsW8MZbb/mOErcwZG7X/x4umT2BrhP++YPnTrzuKq78dBaVa9f0kExEJFiRAkvYLSihK0DSj2/B2Vd05+meN/PEFTdycqdzqH90U9+xSjVxwgR+dcMNvmOUSRgyr5/6JvNuf+AH26s1rE+D9u3IzdzsIZWIiMQjdAVIw2Obs37JcvL27acgEmHV/E9p3aWj71il+uTjj9m5c6fvGGUShszbFi7lwDe7f7C9zT23sfS54eCch1QiIhKPMhcgZja6PILEK3PVGo5r15bUWjWpVLUKrX5+NrUbNfAZSSqQ9E5ns3fLNr75co3vKCIigSkosITdglLqabhmNuX7m4DzzKw2gHPuf8srWEk2r1nPrBde5vZ/PsP+PXv5+stVFEQiQceQCii5ahVa9rmWub+533cUERGJIdY6IE2Bz4ARgKOwAMkAni7tTWbWF+gLcF76cZxSt9GPT3qQDydO58OJ0wG45Hd92Zm1NaH7l3BKa9qY1CaN6DruHwBUa1CfC8YMY/Z1t7N/+w7P6URE5GCxCpAM4HfAn4B7nXOLzGyvc+7d0t7knBsODAe485RzE96Ir163NjnZO6nTqCFtunRiSO9bE30ICaFdq9Yyrct3Z0X9YvrLzO71Gw7s3OUxlYhI+Stw4VuIrNQCxDlXADxjZhOi/26O9Z4g9HnmMdJq1ySSH2HCY0PYu+uHExErkmf++lfOOvts6tSpw7wPP+S5Z55hwrhxvmOVKgyZzxz0IPXbtaFK7VpcNOMVPhs2inWTZviOJSIicTBXhjMFzOxi4Bzn3IPxvqc8RkDK0xu7N/iO8JPweN0TfEcosys/neU7goiER6BDEu///OcJ+13b4b//DSR7mUYznHPTgenllEVEREQOQ0GB7wRlF7p1QERERCT8VICIiIhI4LxPKBUREZEfJxLCs2A0AiIiIiKBUwEiIiIigVMLRkREJOSCvIZLomgERERERAKnAkREREQCpxaMiIhIyOksGBEREZE4qAARERGRuJnZC2a2xcyWHbTtYTP72swWRW8XxdqPWjAiIiIhVxBsC+Yl4G/A6O9tf8Y591S8O9EIiIiIiMTNOTcXyP6x+yn3EZDcgvzyPkRCrXpjiO8IZTav77O+I5RZv82bfEcokzV5udxzzDG+Y5TZuvXrfUcQkZ+O35rZ9cB84G7n3I7SXqwREBERkZCLOEvYzcz6mtn8g25944gwFDgOaAtkAk/HeoPmgIiIiEgR59xwYHgZ37P52/tm9k9gWqz3aAREREREfhQzSz/o4WXAspJe+y2NgIiIiIRcxAV3LDN7BegMHGVmG4H+QGczaws4YB1wS6z9qAARERGRuDnnehazeWRZ96MWjIiIiAROIyAiIiIhF/BCZAmhERAREREJnAoQERERCZxaMCIiIiEXUQtGREREJDYVICIiIhI4tWBERERCLsiFyBJFIyAiIiISuFCOgHS9/hp+fuX/gnNs/HI1L/zpUfIPHPAd6xBrvt7K758eW/R4w+Zs7rzmAjZn7+Kd+SuolJLC0Q3rMuiOK6iZVs1j0u+c8MD91DmnA3k7dvDpdTcUbU+/8grSr7gcF4mw4/0PWPf3oR5TlmzMnJnsyc2lIFJAJJLPbZdd7TtSTJ06daJf//4kJyczbuxYhg6tmF9bEZFEC10BUrtBfbr0vpqHuvckb/9+bh3yKGdd1JX3Jk33He0Qxzapz+QhdwAQiRTQ8ebH6XpWK9Zu2sbdvS8kJTmZwaNn8I/X3uXe67t5Tlto8xv/YdNrEznxoT8Vbat1+mnUO/dcPr3+RlxeHpVq1/aYMLY/9L6RXTt2+o4Rl6SkJAYMHEjvXr3IyspiypQpzJw1i1UrV/qOJiIhE0FnwQQiOTmZylWrkJScTOWqVdm5ZavvSKX6YOlqmjWsS5MGdTi37QmkJCcD0PbEZmRt/8Zzuu/sWryY/F27DtnW6NJL2fDyy7i8PADydobjl3sYtG3blvXr1rFhwwby8vKYOnUqF3bt6juWiEggyjQCYmbnAmcCy5xzb5VPpNLt3LKVN1/8N0/OnkTevv0sf/9jlr//sY8ocZs+bwmX/LzND7a/9vYCfnHOqR4Sxa/a0c2o1aYNzfv2peDAAdb+7XlyPv/cd6xiOecY/NIInHNMfWU808dN8B2pVA0bNWJTZmbR48zMTNqedprHRCIiwSl1BMTMPj7o/s3A34AaQH8zu7+csxUrtWYN2p7fkT92vZy7O19ClWpVad+9YrQwinMgL5+3P1lBtw6nHLJ96KvvkJyUxP92bOspWXwsOZmUGjVY3PcW1j7/d04a+IjvSCW685e9uKXHldz/q1u4tHdPWp/RznekUhU3YOpcCKeyi4h3EZe4W1BitWAqHXS/L9DVOfcIcCHQq6Q3mVlfM5tvZvM/37ElATG/0+rsM9j29SZyduwkkh9hwcw5HN+24o4izP30S04+tjFH1a5RtO31dxYyZ/7nPPX7qzGr2H27A1u2sv3ddwHIWbEC5xwpFXQeyPZoK25ndjbzZs7mpNatPScqXVZWFo3T04sep6ens2XzZo+JRESCE6sASTKzOmZWDzDn3FYA51wukF/Sm5xzw51zGc65jJPqNEhgXNieuZlj25xC5apVAGjZPoNNa9Yl9BiJNP2/i7n43O/aL3MXfsk/X3+XoQ9cR7UqlT0mi8/2//6XWu0KRxKqNmtGUkoK+RVwHkjVatWolpZadD/j3A6sreCTORcvXkzzFi1o2qwZlSpVonv37sycOdN3LBGRQMSaA1ILWEDhaLEzs0bOuSwzq07xI8jlbu2S5Sx46236vTqKgkiEr1Z8ydzxk3xEiWnv/gO8v3gVA269rGjbwBFTOJAX4aZHXgSgzYnNGHDrpb4iHuJnD/en1mmnkVK7Fme8/hpfjXyBzdOmc8KDD3Dav0bh8vL58tG/+I5ZrDpH1WPA3/8KQHJKCrOnTOeTufM8pypdJBKhX79+jB49muTkZMaPH8/KCl40iUjFFPEd4DDY4fSczSwVaOicWxvrtX1atQ9VU3vkhHt9RyizeX2f9R2hzPpt3u47Qpmsycv1HeGwrFu/3ncEkZ+qQP9I/1fbrgn7XXvdopmBZD+sdUCcc3uAmMWHiIiISHFCtxCZiIiIHCqMLZhQLkQmIiIi4aYCRERERAKnFoyIiEjI6VowIiIiInFQASIiIiKBUwtGREQk5CIhvI6URkBEREQkcCpAREREJHBqwYiIiIScFiITERERiYMKEBEREQmcWjAiIiIhF8YWTLkXIGsP7CrvQyRU84v+4DvCT0KXtHTfEcokK2+v7whltocCmh9zjO8YZbZu/XrfEUQkAGrBiIiISODUghEREQm5MLZgNAIiIiIigVMBIiIiIoFTC0ZERCTkIuhaMCIiIiIxqQARERGRwKkFIyIiEnI6C0ZEREQkDipAREREJHBqwYiIiIRcxOksGBEREZGYVICIiIhI4NSCERERCTmdBSMiIiISh1AWIGPmzGTE9EkMnzKRoa+P9x0nLp06dWL2228z5913ue2223zHiSlseQG6Xn8NA6aMYcDkf9N38ABSKlf2HalUjdLTeXHsGKbOnsmUmW/S+6YbfUeKKYzfFyJSMYW2BfOH3jeya8dO3zHikpSUxICBA+ndqxdZWVlMmTKFmbNmsWrlSt/RihW2vAC1G9SnS++reah7T/L27+fWIY9y1kVdeW/SdN/RSpQfyefJRx9jxbLlpKal8eq0qXwwbx6rV67yHa1YYfy+EPmpOOKuBWNmZ5lZzej9amb2iJlNNbMnzKxWMBHDr23btqxft44NGzaQl5fH1KlTubBrV9+xShS2vN9KTk6mctUqJCUnU7lqVXZu2eo7Uqm2bdnKimXLAdiTm8uaVato0LCR51QlC+v3hYhUTLFaMC8Ae6L3nwNqAU9Et71YjrlK5Zxj8EsjGDZpAhf/8ipfMeLWsFEjNmVmFj3OzMykYaOK+4smbHkBdm7Zypsv/psnZ09iyLvT2JuTy/L3P/YdK26Nmzah5cmtWLJoke8oJQrj94WIVFyxWjBJzrn86P0M59zp0fvzzKzEn5Rm1hfoC/Cz+o1oXLPOj096kDt/2YvtW7ZSu25dBo8awYY1a1jyyYKEHiORrJhtrgIvGhO2vACpNWvQ9vyO/LHr5ezdvZvbnvkL7bt348OpM3xHiyk1NZXnhg1l0ICB5Obk+I5TojB+X4j8VBxxLRhgmZndFL2/2MwyAMzsRCCvpDc554Y75zKccxmJLj4AtkeH1ndmZzNv5mxOat064cdIpKysLBqnpxc9Tk9PZ8vmzR4TlS5seQFanX0G277eRM6OnUTyIyyYOYfj257qO1ZMKSkpPDtsKNMmTWbWjDd9xylVGL8vRKTiilWA/BroZGargVbAB2a2Bvhn9LnAVa1WjWppqUX3M87twNoKPglu8eLFNG/RgqbNmlGpUiW6d+/OzJkzfccqUdjyAmzP3MyxbU6hctUqALRsn8GmNev8horDwCefYM2qVYwaMdJ3lJjC+H0h8lMRSeAtKKW2YJxz3wA3mlkN4Njo6zc657z92VPnqHoM+PtfAUhOSWH2lOl8MneerzhxiUQi9OvXj9GjR5OcnMz48eNZWYGLprDlBVi7ZDkL3nqbfq+OoiAS4asVXzJ3/CTfsUp1ekYGPa64nC9WfM7ENwrP1nl28GDmvjPHb7AShPH7QkQSz8xeAC4BtjjnToluqwuMA5oD64CrnXM7St1Pefdwzz++VagaU2vycn1H+EnokpYe+0UVyAe54Ws17KHAd4TDsm79et8RRBKhuGlT5ebPrc9L2O/aR5e8U2p2M+sI5ACjDypAngSynXOPm9n9QB3n3B9L209o1wERERGRQkFeDdc5N9fMmn9vcw+gc/T+KGAOUGoBEsqVUEVERKR8mFlfM5t/0K1vHG9r6JzLBIj+2yDWGzQCIiIiIkWcc8OB4eV9HBUgIiIiIVcB1gHZbGbpzrlMM0sHtsR6g1owIiIi8mNNAW6I3r8BmBzrDSpAREREJG5m9grwAfAzM9toZn2Ax4GuZrYS6Bp9XCq1YEREREIuyBaMc65nCU91Kct+NAIiIiIigVMBIiIiIoFTC0ZERCTkCkJ4ZWqNgIiIiEjgVICIiIhI4NSCERERCbkKsBBZmZV7AaKry5a/e2qf6DtCmf1nb8xF8uRHOrZSmu8IZbYmL5fmxxzjO0aZ6Oq9IodHLRgREREJnFowIiIiIRfGFoxGQERERCRwKkBEREQkcGrBiIiIhFxEC5GJiIiIxKYCRERERAKnFoyIiEjI6SwYERERkTioABEREZHAqQUjIiIScgU6C0ZEREQkNhUgIiIiEji1YEREREJOZ8GIiIiIxCGUIyCdOnWiX//+JCcnM27sWIYOHeo7UkxhyHz+I3fTvONZ7M3eyStX9AWgw+9vpkWn9kTy8vlm4yZm93uKA7tzPSctXlqNGtz5WH+OOfF4cI5nH+jP54uW+I5Vokbp6Qx65mmOql8fV1DA+DGv8PKLL/mOVaoxc2ayJzeXgkgBkUg+t112te9IMYXhsyfyUxS6AiQpKYkBAwfSu1cvsrKymDJlCjNnzWLVypW+o5UoLJk/n/wWS1+ZzAWP3Ve0bSpRNmUAABtnSURBVMOHC/ngryNxkQLOvuvXtOvTkw+eHeExZcn6/vk+Fvz3PQbdeQ8plVKoUrWa70ilyo/k8+Sjj7Fi2XJS09J4ddpUPpg3j9UrV/mOVqo/9L6RXTt2+o4Rl7B89kR+rCOuBWNmd5pZs6DCxKNt27asX7eODRs2kJeXx9SpU7mwa1ffsUoVlsybFi5l367dh2zb8MECXKQAgM1LVlC9wVE+osVULS2NUzLa8daE1wHIz8snd/fuGO/ya9uWraxYthyAPbm5rFm1igYNG3lOdWQJy2dP5Kco1hyQgcBHZvZfM/uNmdUPIlRpGjZqxKbMzKLHmZmZNGxUsX9ohzFzcVpe+j+sf+8T3zGKlX50U77ZsYPfPz6Av04ax52P9adKtYo9AnKwxk2b0PLkVixZtMh3lFI55xj80giGTZrAxb+8ynecmI6Uz57IkShWAbIGaEphIdIO+MzMZpjZDWZWo6Q3mVlfM5tvZvN35+QkMC5YMdtcBV+AJYyZv6/dr6+lIBLhy+mzfUcpVlJyMse3Ook3xkzgzkt/yb49e7mq7698x4pLamoqzw0byqABA8lN8Ocl0e78ZS9u6XEl9//qFi7t3ZPWZ7TzHalUR8JnTyQeBc4l7BaUWAWIc84VOOfecs71ARoDfwe6UViclPSm4c65DOdcRo3q1RMYF7Kysmicnl70OD09nS2bNyf0GIkWxswHO6l7V1p0PIuZDzzuO0qJtmdtZlvWZr5YshSA996cyfEnn+Q5VWwpKSk8O2wo0yZNZtaMN33HiWn7lq0A7MzOZt7M2ZzUurXnRKUL+2dP5EgWqwA55A8I51yec26Kc64ncHT5xSrZ4sWLad6iBU2bNaNSpUp0796dmTNn+ogStzBm/tbRHTI4/aZfMu13/cjft993nBLt2LadrVmbadLiGADanH0WX60qsUauMAY++QRrVq1i1IiRvqPEVLVaNaqlpRbdzzi3A2sr+GTOMH/2RI50sc6C+WVJTzjn9iY4S1wikQj9+vVj9OjRJCcnM378eFZW8B+CYcl84eMP0iSjNVVr1+LGt8bw0dDRtPvVNSRXrkSPYU8AsHnpCuY8+pznpMX7x8DHufepQaRUqkTWxo08e38/35FKdXpGBj2uuJwvVnzOxDemA/Ds4MHMfWeO32AlqHNUPQb8/a8AJKekMHvKdD6ZO89zqtKF5bMn8mOF8SwYK+9+aPNjjgnfVyVk7ql9ou8IZfafvVt8RyiTtft3+Y5QZo0qhWcS7rfW5FXMNWZKs279et8RpGIqbgpSubmmZUbCfteOXTE/kOxaCVVEREQCF7qFyERERORQkRCe3aUREBEREQmcChAREREJnFowIiIiIVcQwrNgNAIiIiIigVMBIiIiIoFTC0ZERCTkdBaMiIiISBxUgIiIiEjg1IIREREJuQK1YERERERiUwEiIiIigSv3FsyS37cr70MkVOtnFviOUGa/HXOr7whlNrHHQ74jlMmAusf5jlBm/bJX+45QZo2Tq/qOUCbvT/0LLH/Nd4yyO/kK3wkkwSJaiExEREQkNhUgIiIiEjidBSMiIhJyBa7Ad4Qy0wiIiIiIBE4FiIiIiAROLRgREZGQK9BZMCIiIiKxqQARERGRwKkFIyIiEnKREF4LRgWIiIiIxM3M1gG7gQiQ75zLOJz9qAARERGRsjrPObftx+xABYiIiEjI6SwYERERCTUz62tm8w+69f3eSxzwlpktKOa5uGkERERERIo454YDw0t5yTnOuU1m1gCYaWafO+fmlvU4KkBERERCriDAs2Ccc5ui/24xs9eBM4EjswCx6vWo9j93Yml1wBWQt3QmBxZNL3q+8uk9qNrxBnYPuwG3b7fHpCXr1KkT/fr3Jzk5mXFjxzJ06FDfkQ6x/0Aevf78Tw7k5RMpKOB/zj6FO6+5oOj5gf+cwsR3FvLpmIf9hYxhzJyZ7MnNpSBSQCSSz22XXe070g+0638P6R3PYn/2TmZedfMhz5143VW0/sMtTDnvcg7s3OUpYckapacz6JmnOap+fVxBAePHvMLLL77kO1ZckpKSeGHya2zdvJl7f32r7ziHOBI+e/LTYWZpQJJzbnf0/oXAgMPZVygKEAoK2Dd3FAVb10ClqqRd+xT5Xy2mIHsjVr0eKce0pmDXVt8pS5SUlMSAgQPp3asXWVlZTJkyhZmzZrFq5Urf0YpUrpTCqEf6kFatCnn5Ea790z/oeNqJtP3Z0SxdtZFde/b5jhiXP/S+kV07dvqOUaL1U99k9bhJnDHwj4dsr9awPg3atyM3c7OnZLHlR/J58tHHWLFsOalpabw6bSofzJvH6pWrfEeL6eqbrmfd6tWkVa/uO8oPHCmfPfnJaAi8bmZQWEOMcc7NOJwdlToJ1cwqm9n1ZnZB9PG1ZvY3M7vdzCodzgEPh9uzo7D4AMjbV1R4AFTt9Cv2/fdfUIFnALdt25b169axYcMG8vLymDp1Khd27eo71iHMjLRqVQDIj0TIzy/AzIhECnhy9H+497punhMeGbYtXMqBb344StfmnttY+txwqMCLCW3bspUVy5YDsCc3lzWrVtGgYSPPqWKr36ghHc7rzNRxr/qOUix99iQRChJ4K41zbo1zrk30drJz7rHDzRxrBOTF6GtSzewGoDowEehCYc/nhsM98OGymvVJrt+CSNaXpBx7BgU52ynYti7oGGXSsFEjNmVmFj3OzMyk7WmneUxUvEikgMvvfZ6vsrZzbbf2tDmxGaOmvUeXM1rSoG5N3/Fics4x+KUROOeY+sp4po+b4DtSXNI7nc3eLdv45ss1vqPErXHTJrQ8uRVLFi3yHSWmux56kOcfH0xqWprvKCUK+2dP5HDEKkBOdc61NrMU4GugsXMuYmYvA4tLelP0tJy+AM9e1ZabOrRITNpKVUm9+D72vfsCFESocuYV5E48rNZToKyYba4C/qWbnJzE5CF3sCt3L7c/8TKfLF/LjPeX8a+Bv/YdLS53/rIX27dspXbdugweNYINa9aw5JMFvmOVKrlqFVr2uZa5v7nfd5S4paam8tywoQwaMJDcnBzfcUrV4fzO7NiezRfLlnPaWWf6jlOisH/2RA5HrHVAksysMlADSAVqRbdXAUpswTjnhjvnMpxzGQkrPpKSSb3kXvI+n0v+6o9IqtUIq9mQ6r2HUP1Xw7Dq9Ui79ikstXZijpdAWVlZNE5PL3qcnp7Ols0Vt9dfM60aZ518LB8tW8NXWdu58DdPc/4tT7J3fx5df/OU73gl2r6lcB7Qzuxs5s2czUmtW3tOFFta08akNmlE13H/4BfTX6Zag/pcMGYYVerV8R2tWCkpKTw7bCjTJk1m1ow3fceJqXW70zm3y/m8Nnc2A/46hHZnt6f/kMG+Y5UorJ898a/AuYTdghJrBGQk8DmQDPwJmGBma4D2wNhyznaIqhfcTiT7aw58OhWAgu1fkTP8pqLnq/9qGLlj7q2QZ8EsXryY5i1a0LRZMzZnZdG9e3fuvPNO37EOkf1NDikpydRMq8a+/Xm8v2QVN1/WifdeeLDoNadd+zAz/36Px5Qlq1qtGpZk7M3dQ9Vq1cg4twOj/1axzjQqzq5Va5nW5aqix7+Y/jKze/2mQp4FAzDwySdYs2oVo0aM9B0lLsMGD2HY4CEAnHbWmVx786945A/3ek51qLB/9kQOV6kFiHPuGTMbF72/ycxGAxcA/3TOfRxEQIDkxidRuVVnIlvXkdLraQD2v/dv8tctDCrCjxKJROjXrx+jR48mOTmZ8ePHs7ICnQEDsGXHbu7/v1eJFDhcQQHdzjmV8zJO8h0rbnWOqseAv/8VgOSUFGZPmc4nc+d5TvVDZw56kPrt2lCldi0umvEKnw0bxbpJhzWBPHCnZ2TQ44rL+WLF50x8o/A0+GcHD2buO3P8Bgu5sH/2RA6XlfdchF3PXl7xJjuUovUzFXvOQHHWvTHEd4QyO7/HQ74jlMlvajT2HaHM+mWv9h2hzGonV/YdoUzen/oX3xEOz8lX+E7wU1Dc9L9y0+HYnyXsd+37a74IJLuuBSMiIiKBUwEiIiIigQvHSqgiIiJSoiDPXkkUjYCIiIhI4FSAiIiISODUghEREQm5ggp8PbSSaAREREREAqcCRERERAKnFoyIiEjIqQUjIiIiEgcVICIiIhI4tWBERERCriB8HRiNgIiIiEjwyn0EJLVNx/I+REJ1SdvkO0KZZf39cd8Rymz6gAt8RyiTlg9M9h2hzHrWbOE7QpkNGnuH7whlsuCOp31HKLO1O1KBob5jlMmVn87yHUHKgVowIiIiIaezYERERETioAJEREREAqcWjIiISMipBSMiIiISBxUgIiIiEji1YERERELOha8DoxEQERERCZ4KEBEREQmcWjAiIiIhp7NgREREROKgAkREREQCpxaMiIhIyIWvAaMREBEREfFABYiIiIgELhQtmMzsHB54aTbbd+3BzLjq3FZc16U1KzZsY8CYd9mfFyElKYk/9/w5rVs09B23WF2vv4afX/m/4Bwbv1zNC396lPwDB3zHKpJUuyG1bniY5Jr1cM6xd97r7JkzliqndaH6xX1Jadic7YNvJP+rFb6jFsn6Zi9/nrSU7TkHMIMrTm9Gr/bHAPDKR+sZ+8lXJCcZPz+hPr/v+jPPaYvXqVMn+vXvT3JyMuPGjmXo0KG+I/3AFQPu46SOZ5OTvZPnLr8JgGo1a9Dzqf7UadyIHZuyGHPPw+zbleM5aaE1X2/l90+PLXq8YXM2d15zAQ3r1eRv42azeuNWJjxxG6ce39RjykMdc++D1Gp/Dvk7d/BZn94AtHhoAFWbHQ1AcvUaRHJ2s6LvjR5Tfqdd/3tI73gW+7N3MvOqmw957sTrrqL1H25hynmXc2DnLk8Jf3rCeBZMKAqQlGTjvis70Oro+uTuO8BVf3mVs1s2ZcjED/jNxRn8/JRjmLt0PUMmfshLd/fwHfcHajeoT5feV/NQ957k7d/PrUMe5ayLuvLepOm+o32nIJ/dE58lf8MXWJVU6v1xNPs//4j8TavZOfw+avV8wHfCH0hOSuLuC0+iZXpNcvfn03P4B7Q/rh7ZOfuZ88UWJtx6DpVTksjO3e87arGSkpIYMHAgvXv1IisriylTpjBz1ixWrVzpO9ohFkyewQevvM5Vjz1YtK1Tn2tZ/dFC3h05hk59rqVzn2uZ8cxwjym/c2yT+kwecgcAkUgBHW9+nK5ntWLfgTz+775e9B82yXPCH9r+5htsmfQqLe7vV7Rt7cDv7je99Q4iuRWjwANYP/VNVo+bxBkD/3jI9moN69OgfTtyMzd7SiZhErMFY2bHmdk9ZvacmT1tZreaWa0gwn2rfq00Wh1dH4C0qpU5tlEdtuzMBTNy9uUBsHvfAerXTg0yVpkkJydTuWoVkpKTqVy1Kju3bPUd6RAFu7aTv+ELANz+PeRvXkdy7fpENq8jsmW953TFq1+jCi3TawKQViWFY+unsWXXPsbP38BN5x5L5ZTCb++6aVV8xixR27ZtWb9uHRs2bCAvL4+pU6dyYdeuvmP9wLoFS9jzze5DtrU67xwWTp4BwMLJM2h13rk+osX0wdLVNGtYlyYN6nBc0wYc26S+70jFylmyiMiukkcL6nQ+n+y3ZwaYqHTbFi7lwPe+JwDa3HMbS58bHs51wUPOJfAWlFILEDO7ExgGVAXOAKoBzYAPzKxzuacrxtfbdrFiwzZat2jI/Vedw1OvfUCXB0bz1Ksf8PtL2/uIFNPOLVt588V/8+TsSQx5dxp7c3JZ/v7HvmOVKLluOpWa/oy8dct9R4nb1zv38nnmbk5tWpv12/ewcP0Oeo/4kD4vfcyyr7/xHa9YDRs1YlNmZtHjzMxMGjZq5DFR/KrXq8vubdkA7N6WTfV6dTwnKt70eUu45OdtfMf4Uaq3bkvejmz2f73Rd5RSpXc6m71btvHNl2t8R5GQiDUCcjPQzTn3KHAB0Mo59yegG/BMSW8ys75mNt/M5v9z2vsJC5u7L4+7hr/J/VefQ/VqlRk3dzl/vKoDswddzx+v6sBD/3onYcdKpNSaNWh7fkf+2PVy7u58CVWqVaV9926+YxXLqlSj9s1PsOvVIbh9ub7jxGXPgXzuGb+Ie7udRPUqKUQKHLv35fGvPmdxV9cTue/VxbgK+BeZFbOtIuYMqwN5+bz9yQq6dTjFd5Qfpe75F5D99izfMUqVXLUKLftcy/Kho3xHkRCJ5yyYb+eJVAFqADjnvgIqlfQG59xw51yGcy7j5ks6/PiUQF4kwl3D3+TiM0+k62nHAjD5gy+K7v9Pu+NYum5LQo6VaK3OPoNtX28iZ8dOIvkRFsycw/FtT/Ud64eSkqn96yfY+8kM9i+umMXc9+VFCrh7/CIuOjWdLi0LJyA3rFmF81s2xMw4tUltkgx27MnznPSHsrKyaJyeXvQ4PT2dLZvD0TvP2Z5NjaPqAlDjqLrkbN/hOdEPzf30S04+tjFH1a7hO8rhS0qm9rmd2fFOxS5A0po2JrVJI7qO+we/mP4y1RrU54Ixw6hSQUfGjkRHXAsGGAF8YmbDgQ+AvwGYWX0gu5yzFXHO0W/0HI5tVJsbL/huOLVB7VQ++XITAB998TXHNAh0akrctmdu5tg2p1C5auFchJbtM9i0Zp3fUMWo1fsh8rPWseftMb6jxMU5xyNTltPiqDSuO7t50fbzTmrIJ2u3A7B+ey55EUed1BLrZW8WL15M8xYtaNqsGZUqVaJ79+7MnFlx+vylWTHnfU7vUTiKd3qPbnz2znueE/3Q9P8u5uJzw91+qdkug30b1pO3rWLNGfu+XavWMq3LVfzn4t785+Le7N2ylVnX3sr+CliYSsVR6lkwzrnnzGwW0BIY4pz7PLp9K9AxgHwALFydxZSPvuTEJnW5/NHxANzV4ywe7t2Zx8fPIz/iqFIpmYd7dQ4qUpmsXbKcBW+9Tb9XR1EQifDVii+ZO75izcSvdFwbqp11MXlfr6TeA/8GYPeU57GUytS86h6Sqtehzm3PkL/xS3Y8f6fntIUWbdjJtCWbOKFBda4eVtjqu6PLCVx6WhP6T17GFX9/j0rJxsBLT8GsuIaHX5FIhH79+jF69GiSk5MZP348KyvYGTAA1zzxEC3OaEta7VrcP2sCs55/kXdHjqHnU/3JuOwidmZuZszdD/uOeYi9+w/w/uJVDLj1sqJtMz9czsARU8nelcstj42iZYvGjOx3k8eU32nx50eo0eY0UmrV5tRxk9j00gi2/2cadc67oEJNPv3WmYMepH67NlSpXYuLZrzCZ8NGsW7SDN+xJGSsvHvO+e88G6qm9i23j439ogrmsfMiviOUWa1zzvYdoUxaPjDZd4Qy61mzhe8IZTZo7B2+I5TJgjue9h2hzNbuqLhnC5bkyk8rdguqBIH+1XP8Mcck7HftqvXrA8mulVBFREQkcCpAREREJHChWAlVREREShaquQ5RGgERERGRwKkAERERkcCpBSMiIhJyasGIiIiIxEEFiIiIiAROLRgREZGQUwtGREREJA4qQERERCRwKkBERERCziXwFouZdTOzL8xslZndf7iZVYCIiIhIXMwsGXge+AXQCuhpZq0OZ18qQERERCReZwKrnHNrnHMHgLFAj8PZkTkXxrmzhcysr3NuuO8c8QpbXghf5rDlBWUOQtjygjIHIWx5g2JmfYG+B20a/u3XycyuBLo5534dfXwdcJZz7rdlPU7YR0D6xn5JhRK2vBC+zGHLC8ochLDlBWUOQtjyBsI5N9w5l3HQ7eAizYp7y+EcJ+wFiIiIiARnI9DsoMdNgU2HsyMVICIiIhKvT4ATzKyFmVUGrgGmHM6Owr4Sath6d2HLC+HLHLa8oMxBCFteUOYghC2vd865fDP7LfAmkAy84Jxbfjj7CvUkVBEREQkntWBEREQkcCpAREREJHChLEAStQxsUMzsBTPbYmbLfGeJh5k1M7N3zGyFmS03s9/5zhSLmVU1s4/NbHE08yO+M8XDzJLN7FMzm+Y7SzzMbJ2ZLTWzRWY233eeeJhZbTN71cw+j35Pn+07U2nM7GfRr++3t11mdpfvXKUxs99HP3fLzOwVM6vqO1MsZva7aN7lFf3re6QK3RyQ6DKwXwJdKTwd6BOgp3PuM6/BSmFmHYEcYLRz7hTfeWIxs3Qg3Tm30MxqAAuASyv419iANOdcjplVAuYBv3POfeg5WqnM7A9ABlDTOXeJ7zyxmNk6IMM5t813lniZ2Sjgv865EdFZ+6nOuZ2+c8Uj+vPuawoXelrvO09xzKwJhZ+3Vs65vWY2HnjDOfeS32QlM7NTKFzB80zgADADuM05t9JrsJ+YMI6AJGwZ2KA45+YC2b5zxMs5l+mcWxi9vxtYATTxm6p0rlBO9GGl6K1CV9dm1hS4GBjhO8uRysxqAh2BkQDOuQNhKT6iugCrK2rxcZAUoJqZpQCpHOa6EAFqCXzonNvjnMsH3gUu85zpJyeMBUgTYMNBjzdSwX85hpmZNQdOAz7ymyS2aDtjEbAFmOmcq+iZnwXuAwp8BykDB7xlZguiyzVXdMcCW4EXo62uEWaW5jtUGVwDvOI7RGmcc18DTwFfAZnAN865t/ymimkZ0NHM6plZKnARhy6uJQEIYwGSsGVgpXRmVh14DbjLObfLd55YnHMR51xbClfmOzM6zFohmdklwBbn3ALfWcroHOfc6RReCfP2aHuxIksBTgeGOudOA3KBCj9vDCDaLvpfYILvLKUxszoUjkK3ABoDaWbW22+q0jnnVgBPADMpbL8sBvK9hvoJCmMBkrBlYKVk0XkUrwH/ds5N9J2nLKJD7HOAbp6jlOYc4H+jcyrGAueb2ct+I8XmnNsU/XcL8DqFLdGKbCOw8aDRsFcpLEjC4BfAQufcZt9BYrgAWOuc2+qcywMmAh08Z4rJOTfSOXe6c64jhS1yzf8IWBgLkIQtAyvFi07oHAmscM4N8Z0nHmZW38xqR+9Xo/CH4ud+U5XMOfeAc66pc645hd/DbzvnKvRfjWaWFp2UTLSNcSGFQ9kVlnMuC9hgZj+LbuoCVNjJ1N/Tkwrefon6CmhvZqnRnx1dKJw3VqGZWYPov0cDlxOOr/URJXRLsSdyGdigmNkrQGfgKDPbCPR3zo30m6pU5wDXAUujcyoAHnTOveExUyzpwKjoWQNJwHjnXChObQ2RhsDrhb9jSAHGOOdm+I0UlzuAf0f/YFkD3OQ5T0zReQldgVt8Z4nFOfeRmb0KLKSwjfEp4Vji/DUzqwfkAbc753b4DvRTE7rTcEVERCT8wtiCERERkZBTASIiIiKBUwEiIiIigVMBIiIiIoFTASIiIiKBUwEiIiIigVMBIiIiIoH7f6L0Kb06LUDyAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.multiclass import OneVsRestClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "gnb=GaussianNB()\n",
    "gnb.fit(X_train,y_train)\n",
    "model_assess(gnb,\"Simple Gaussian NB\")\n",
    "clf = OneVsRestClassifier(GaussianNB()).fit(X_train,y_train)\n",
    "model_assess(clf,\"One Vs Rest\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "d4wB41XIDJJt"
   },
   "source": [
    "## Grid Search Gaussian Naive Bayes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "v4YfvqWWwgiA",
    "outputId": "7340fd7b-3e34-4a72-922f-3ca90ca70032"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 100 candidates, totalling 500 fits\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'var_smoothing': 8.111308307896872e-05}"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nb_classifier = GaussianNB()\n",
    "\n",
    "params_NB = {'var_smoothing': np.logspace(0,-9, num=100)}\n",
    "gs_NB = GridSearchCV(estimator=nb_classifier, \n",
    "                 param_grid=params_NB,    # use any cross validation technique \n",
    "                 verbose=1, \n",
    "                 scoring='accuracy') \n",
    "gs_NB.fit(X_train, y_train)\n",
    "\n",
    "gs_NB.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Uo3nffUsggiQ",
    "outputId": "3f1b1d7f-1901-40cf-87bd-5ee22f837559"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "GaussianNB GridSearchCV :\n",
      "Accuracy score: 0.46896896896896895\n",
      "Overall Precision: 0.4606482379884576\n",
      "Overall Recall:0.4755693757933862\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.39      0.46      0.42       191\n",
      "           1       0.68      0.88      0.77       214\n",
      "           2       0.34      0.29      0.31       193\n",
      "           3       0.39      0.27      0.32       202\n",
      "           4       0.65      0.30      0.41       201\n",
      "           5       0.49      0.39      0.43       202\n",
      "           6       0.45      0.87      0.59       187\n",
      "           7       0.45      0.76      0.57       177\n",
      "           8       0.52      0.41      0.46       211\n",
      "           9       0.26      0.12      0.17       220\n",
      "\n",
      "    accuracy                           0.47      1998\n",
      "   macro avg       0.46      0.48      0.44      1998\n",
      "weighted avg       0.46      0.47      0.44      1998\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiAAAAGfCAYAAABm/WkhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3dd3hUZfr/8fedBARCb0kQpNgAlRoLRUAQ1+8qimLXtawa159l1dVdywqr6FpYQV0VREFhXURAlGYDVlAUVJCiiEpAEISEXkKRZOb5/ZExC24yBSfn5ODndV1zMXNm5pwPXDPJzXM/5znmnENERETESyl+BxAREZFfHxUgIiIi4jkVICIiIuI5FSAiIiLiORUgIiIi4jkVICIiIuI5FSAiIiISFzOrYmafmtliM1tqZg9Etjc3s0/MbLmZvWZmlWPtSwWIiIiIxOtHoKdzri3QDjjTzE4BHgOGOOeOBrYC18bakQoQERERiYsrVhB5WClyc0BPYEJk+yigb6x9pZVLwgNpqVUREfm1MU+PtvT1pP2uteMvuAHI2W/TcOfc8JLnzVKBBcBRwLPACmCbc64o8pK1wOGxjuNFASIiIiIBESk2hkd5PgS0M7PawBtAq9JeFus4asGIiIhIwpxz24BZwClAbTP7aVCjMbAu1vtVgIiIiEhczKxBZOQDM6sKnA4sA94HLoi87CpgUqx9qQUjIiIScC4UStq+YkxeyQJGReaBpADjnHNTzewrYKyZPQQsBEbEOo4KEBEREYmLc24J0L6U7SuBkxLZl1owIiIi4jmNgIiIiARdqCj2ayoYjYCIiIiI51SAiIiIiOfUghEREQk4F05eC8arJVw1AiIiIiKeUwEiIiIinlMLRkREJOiSuBCZVzQCIiIiIp5TASIiIiKeUwtGREQk4FwAFyKLWYCYWUvgXOBwwFF8id3Jzrll5ZxNREREDlFRWzBm9hdgLMWnBX8KfBa5/6qZ3V3+8URERORQFGsE5FrgOOdc4f4bzWwwsBR4tLQ3mVkOkAPw/PPPk5OTk4SoIiIiUqpDsAUTBhoBq3+2PSvyXKmcc8OB4T89POh0IiIickiKVYDcBsw0s+XAmsi2I4CjgJvLM5iIiIgcuqIWIM65d8zsGOAkiiehGrAW+Mw5F7xVT0RERA5BybwWjFdingXjnAsD8zzIIiIiIr8SWohMREREPKeFyERERIJO14IRERERiU0FiIiIiHhOLRgREZGAC+K1YDQCIiIiIp5TASIiIiKeUwtGREQk6NSCEREREYlNIyAiIiIB58JaB0REREQkJhUgIiIi4jm1YERERAJO64CIiIiIxKHcR0BaN21e3odIuq9Wf+d3BBERkUOaWjAiIiJBpxaMiIiISGwqQERERMRzasGIiIgEnBYiExEREYmDChARERHxnFowIiIiQaezYERERERiUwEiIiIinlMLRkREJOB0LRgRERGROKgAEREREc+pBSMiIhJ0asGIiIiIxBa4EZDMrCweGfIE9Rs0wIXDjBvzKq+89LLfsURERCQBgStAikJFPP7Qwyz7cinV0tOZMHUKc+fMYcXyXL+jiYiI+ELXgvHApg0bWfblUgB279rFytxcGmZk+pxKREREEhG4AmR/jRofTqvjWrNk0SK/o4iIiEgCDroAMbNrojyXY2bzzWz+1oKdB3uIqKpVq8ZTw4byyIMD2VVQUC7HEBERCYRQUfJuHvklIyAPlPWEc264cy7bOZddp3qNX3CI0qWlpfHksKFMfXMSM955N+n7FxERkfIVdRKqmS0p6ykgI/lx4jPw8cdYmZvLqBdH+BVBREREfoFYZ8FkAL8Btv5suwEfl0uiGDpkZ3Nuv/P5ZtnXTHxrGgBPDhrEB+/P8iOOiIiI71woeGfBxCpApgLVnXP/M8vTzGaVS6IYPp8/n9ZNm/txaBERkV81M2sCjAYygTAw3Dn3lJn9Dbge2Bh56b3Oubei7StqAeKcuzbKc5clElpEREQCrwj4k3PuczOrASwws+mR54Y45/4R744CtxCZiIiIHMh5dPaKc249sD5yf6eZLQMOP5h9BXodEBEREUmu/ZfSiNxyynhdM6A98Elk081mtsTMRppZnVjHUQEiIiIiJfZfSiNyG/7z15hZdeB14Dbn3A5gKHAk0I7iEZInYh1HLRgREZGgC3u3gJiZVaK4+Pi3c24igHMuf7/nX6D4JJaoNAIiIiIicTEzA0YAy5xzg/fbnrXfy84Dvoy1L42AiIiISLy6AL8DvjCzn5bouBe41MzaAQ5YBdwQa0cqQERERALOq4XInHNzKF6M9OeirvlRGrVgRERExHMqQERERMRzasGIiIgEXQCvBaMREBEREfGcChARERHxnFowIiIiAefVtWCSqdwLkK/eivvCeBXCrRcP4dbju/odI2FPfznH7wgiIiJxUwtGREREPKcWjIiISNDpLBgRERGR2FSAiIiIiOfUghEREQk4r64Fk0waARERERHPqQARERERz6kFIyIiEnAurBaMiIiISEwqQERERMRzasGIiIgEnc6CEREREYlNBYiIiIh4Ti0YERGRgAviQmSBKUB63vA46VUPIyUlhdTUFCYOuom3P/6CZ16byYq1Gxn/2I2ccFRjv2OWqfsVF9KpXx/MjLkTJjPrlfF+RxIREfFNYAoQgFEPXkfdmuklj485IoN//vlyBgx708dUsWUd1ZxO/frwxKXXEyos4sZhT7D0g7ls/H6t39FERER8Eeg5IEc2bkiLwxv4HSOmjBbNWL1kKYV7fyQcCpE7fyFtenXzO5aIiBwiXCictJtXYhYgZtbSzHqZWfWfbT+z/GKVGoRrH3iJ8+98htfe+9TTQ/9S63NXcmTHdlSrVZNKVQ6j9amdqJ3Z0O9YIiIivonagjGzW4GbgGXACDP7o3NuUuTpvwPvlPG+HCAH4PkBN5BzYe9fHPTVv99ARt2abN5WwDUPjKTF4Q048bjmv3i/XshfuZoZI1/hpheG8OPuPfzwbS7hAE4YEhERSZZYc0CuBzo65wrMrBkwwcyaOeeeAqysNznnhgPDAVj6uktG0Iy6NQGoV7s6vU9uzZLlawNTgADMmziNeROnAXD2H3PYlrfR50QiInLI8LB1kiyxWjCpzrkCAOfcKqAH8H9mNpgoBUiy7d67j4I9P5bc/2hxLkcfkeHV4ZOiet3aANTJzKBtr+4seHuGz4lERET8E2sEJM/M2jnnFgFERkLOBkYCJ5R7uojN2wq46bFXAAiFw5x9alu6dTiG6fOWMvDFKWzZsYsbHh5Fq+aNGNH/Gq9iJeTaIQ+TXrsmoaIQ4x8ezJ4dO/2OJCIi4htzruwOiZk1Boqcc3mlPNfFOfdRzCMkqQXjlVsvHuJ3hIPy9Jdz/I4gIiL/5VmXAGDzPV2T9ru23iNzPMkedQTEOVfmQhVxFR8iIiIipQj0OiAiIiISTIFaCVVERET+lwsFarYDoBEQERER8YEKEBEREfGcWjAiIiIB5+U1XJJFIyAiIiLiORUgIiIi4jm1YERERAJOLRgRERGROKgAEREREc+pBSMiIhJwLqyFyERERERiUgEiIiIiniv3Fsw9l/yzvA+RVDN2/uB3hITVTq1M5xbH+h0jYR+v/MbvCCIihwRdC0ZEREQkDipARERExHM6C0ZERCTgXMjvBInTCIiIiIh4TgWIiIiIeE4tGBERkYDTWTAiIiIicVABIiIiIp5TC0ZERCTgwmG/EyROIyAiIiISFzNrYmbvm9kyM1tqZn+MbK9rZtPNbHnkzzqx9qUCREREROJVBPzJOdcKOAW4ycxaA3cDM51zRwMzI4+jUgtGREQk4LxaiMw5tx5YH7m/08yWAYcD5wI9Ii8bBcwC/hJtXxoBERERkRJmlmNm8/e75ZTxumZAe+ATICNSnPxUpDSMdRyNgIiIiEgJ59xwYHi015hZdeB14Dbn3A4zS/g4gShA+j34Z1p260TBlm08df41AFStWYNL/zGAOo0y2boujzF3/o29Owp8Tlq6zKwsHhnyBPUbNMCFw4wb8yqvvPSy37FiSklJYeSk19mYn89d1/3B7zgiIlIGL68FY2aVKC4+/u2cmxjZnG9mWc659WaWBWyItZ9AtGAWTHqHl2788wHbul97GSs++Zwnzr6CFZ98To9rL/MpXWxFoSIef+hh+vTqzSV9z+eyK6/kyKOP8jtWTBddcyWrVqzwO4aIiFQQVjzUMQJY5pwbvN9Tk4GrIvevAibF2lcgCpBVC5awe/vOA7a1Pq0Ln096B4DPJ71D69O6+hEtLps2bGTZl0sB2L1rFytzc2mYkelzqugaZGbQ+bQeTHltgt9RRESk4ugC/A7oaWaLIrffAo8Cvc1sOdA78jiqmC0YMzsJcM65zyKn2pwJfO2ce+sX/RV+oer16rJz0xYAdm7aQvV6MU85rhAaNT6cVse1ZsmiRX5Hieq2++/l2UcHUS093e8oIiISg1cLkTnn5gBlTfjolci+oo6AmNkA4GlgqJk9AjwDVAfuNrP7oryvZAbtoi3rEslzSKtWrRpPDRvKIw8OZFdBxZyvAtC5Zw+2bt7CN5FRGxERkWSLNQJyAdAOOAzIAxpHZrsOovi0m4dLe9P+M2jvOaFHuVyir2DzFmrULx4FqVG/LgWbt5bHYZImLS2NJ4cNZeqbk5jxzrt+x4mqTccOdO3Vk049ulH5sMNIr16dAYMH8cAdd/kdTUREDhGx5oAUOedCzrndwArn3A4A59wewNeV55fN+pgO554JQIdzz+Sr9z/yM05MAx9/jJW5uYx6cYTfUWIaNmgwfbt0p1+3XvS/9Q4WzJ2n4kNEpAJzoeTdvBKrANlnZtUi9zv+tNHMauFhAXLJY/dz4yvP0qBZE+6eMZ7s837L7BFjOOqUjvxp6iscdUpHZo8Y41WchHXIzubcfudzcufOTHxrGhPfmka303r4HUtERMQ35lzZHRIzO8w592Mp2+sDWc65L2IdoLxaMOVl0o7VfkdIWO3Uyn5HOCgfr/zG7wgiIuUl8ZW5foHvLslO2u/a5mPne5I96hyQ0oqPyPZNwKZySSQiIiIJCYc9rXeSIhDrgIiIiMihRQWIiIiIeC4Q14IRERGRsnm1EFkyaQREREREPKcREBERkYDzcv2OZNEIiIiIiHhOBYiIiIh4Ti0YERGRgNM6ICIiIiJxUAEiIiIinlMLRkREJODCOgtGREREJDYVICIiIuK5cm/B7HJF5X2IpNpH8Naz3RDa63eEhL3c+AjmdDnV7xgJ6frRh35HEBEplc6CEREREYmDChARERHxnM6CERERCTinFoyIiIhIbCpARERExHNqwYiIiARcOHgncGoERERERLynAkREREQ8pxaMiIhIwGkhMhEREZE4qAARERERz6kFIyIiEnBqwYiIiIjEQQWIiIiIeC6QLZjuV1xIp359MDPmTpjMrFfG+x0pqkcGDaJnz55s3ryZ355xht9x4hKEzEffczd1unSmcOtWFv7uqpLtWRf0I6vf+bhQiK0fz2XVc0N9TCkiUv5CasGUv6yjmtOpXx+euPR6Hut3Ncd170KDIxr7HSuqiePH8/urror9wgokCJnz33qbpXfcecC2Wh3aU69rVxZeeTULr7iSH8a86lM6ERGJJnAFSEaLZqxespTCvT8SDoXInb+QNr26+R0rqs8+/ZRt27b5HSMhQci8Y/FiinbsOGBbZt++rHnlFVxhIQCFFfzvICLya5VwAWJmo8sjSLzW567kyI7tqFarJpWqHEbrUztRO7Ohn5GkAql6RBNqtW1L2+HPc8Iz/6R6y5Z+RxIRKXfhsCXt5pWoc0DMbPLPNwGnmVltAOfcOeUVrCz5K1czY+Qr3PTCEH7cvYcfvs0lHAp5HUMqKEtNJa1GDRbn3ED1Vq1oOfAB5l94sd+xRETkZ2JNQm0MfAW8CDiKC5Bs4IlobzKzHCAH4LSsIzm+buYvT7qfeROnMW/iNADO/mMO2/I2JnX/Elz7Nmxk8+zZABQsW4ZzjrTatSlSK0ZEpEKJ1YLJBhYA9wHbnXOzgD3OudnOudllvck5N9w5l+2cy0528QFQvW5tAOpkZtC2V3cWvD0j6ceQYNr84YfU6tgRgCpNmpCSlqbiQ0QOeWFnSbt5JeoIiHMuDAwxs/GRP/NjvccL1w55mPTaNQkVhRj/8GD27Njpd6Sohjz9NCd36kSdOnWYM28eTw0ZwvjXXvM7VlRByHzs3wZQq3170mrX4sQ3Xuf7ESPJnzqNo++9h/b/GoUrLOLbh/7ud0wRESmFOefif7HZWUAX59y98b7n1uO7xn+ACuCtnWv8jvCr8HLjI/yOkLCuH33odwQRCQ5PF+b4+NRTk/a7tvOHH3qSPaHRDOfcNGBaOWURERGRgxAO+50gcYFbB0RERESCTwWIiIiIeM73CaUiIiLyy4Q8PHslWTQCIiIiIp5TASIiIiKeUwtGREQk4Ly8hkuyaAREREREPKcCRERERDynFoyIiEjA6SwYERERkTioABEREZG4mdlIM9tgZl/ut+1vZvaDmS2K3H4baz9qwYiIiARc2NsWzMvAM8Don20f4pz7R7w70QiIiIiIxM059wGw5Zfup9xHQNYX7SnvQyRV7luD/Y6QsAW3POF3hITd9f1mvyMkZGXhLmja1O8YCVu1erXfEUTk1+NmM7sSmA/8yTm3NdqLNQIiIiIScCFnSbuZWY6Zzd/vlhNHhKHAkUA7YD0Q83/GmgMiIiIiJZxzw4HhCb4n/6f7ZvYCMDXWezQCIiIiIr+ImWXt9/A84MuyXvsTjYCIiIgEXMh5dywzexXoAdQ3s7XAAKCHmbUDHLAKuCHWflSAiIiISNycc5eWsnlEovtRC0ZEREQ8pxEQERGRgPN4IbKk0AiIiIiIeE4FiIiIiHhOLRgREZGAC6kFIyIiIhKbChARERHxnFowIiIiAeflQmTJohEQERER8VzgRkAaNW/K7YMfKXncsMnhvPb0MN4a/aqPqUq3Y9ce/vrsRL5dk49h/P3mflStXIkBz7/J7r37OLxhHf5x20VUr1bF76gANL3rXmqd0oWibVv56torAGh+/4NUaXIEAKnVaxAq2MmynKt9TFm2MbOms3vXLsKhMKFQETeed5HfkWLq3r07/QcMIDU1ldfGjmXo0KF+RxIR8UTgCpB1363mrvMuAyAlJYXnZ7/NpzPe9zlV6R4eMZVT2x/D03++nH2FRezdV8g1fxvJX67+P046rgUTZs7nxTc/5LbLevsdFYDN777Fhjcn0Pzu/iXbvhv43/uN/3ALoV0FfkSL2x1XXM2Ordv8jhGXlJQUHhw4kCsuv5y8vDwmT57M9BkzyF2+3O9oIhIwIXQWjKeO73QSeWvWsmldnt9R/kfB7r189tUqLjg9G4DKldKomV6V79Zt4sTWzQHo0vYo3psX84KBnilYsojQjh1lPl+nR0+2/Ge6h4kObe3atWP1qlWsWbOGwsJCpkyZwhm9K0YxKiJS3hIqQMysq5ndYWZnlFegRHT57Rl8NO1dv2OUak3+FurWTOeeZ16n75/+yX3PTmT33n0cc0QGMz9bBsA7H3/J+k3bfU4an+pt2lG4dQs//rDW7yhlcs4x6OUXGfbmeM66+EK/48SUkZnJuvXrSx6vX7+ejMxMHxOJiHgnagFiZp/ud/964BmgBjDAzO4u52xRpVVKI7tnd+a+M8PPGGUqCoX5auU6Lv3Nybz5xC1UrVKJ4RNn8/BN5zPm7Xmcf+cz7NrzI5XTUv2OGpe6PU9ny38q5r/1T269+HJuOPcC7v79DfS94lLanNjR70hRlTZg6lwAp7KLiO9CLnk3r8QaAam03/0coLdz7gHgDODyst5kZjlmNt/M5q/ctikJMf9Xu1O78N1XX7N985Zy2f8vlVmvFpn1atL2mCYAnNnpeL5auY4jGzdk5IDfM/EfN3PWqW1pklnP56RxSEmldtcebH2/YhcgmzdsBGDbli3MmT6Tlm3a+Jwoury8PBplZZU8zsrKYkN+vo+JRES8E6sASTGzOmZWDzDn3EYA59wuoKisNznnhjvnsp1z2S1q109i3P/qetZvmDPtnXLZdzI0qFODzPq1WPlD8S/FuUtWcGSThmzeVjyJMxwOM3T8+1zym5P8jBmXmh2z2btmNYWbNvodpUxVqlalanq1kvvZXTvzXQWfzLl48WKaNW9O4yZNqFSpEn369GH6dM2xEZFfh1hnwdQCFlA8WuzMLNM5l2dm1Sl9BNkTlatUoU2Xkxk+4O9+RYjL/df14c4nx1FYFKJJRh0eufkC3pz1OWPengdA71OOo1/PitMmaP7XB6jRtj1ptWpzwmtvsu7lF9n89lTqnHZ6hZ98Wqd+PR587mkAUtPSmDl5Gp99MMfnVNGFQiH69+/P6NGjSU1NZdy4cSyv4EWTiFRMIb8DHAQ7mJ6zmVUDMpxz38V67YUtOwaqqT3+9Xv9jpCwBbc84XeEhN31fTBOlf3JysJdfkc4KKtWr/Y7gsivlaf/Sf9Xu95J+137u0XTPcl+UOuAOOd2AzGLDxEREZHSBG4hMhERETlQEFswgV6ITERERIJJBYiIiIh4Ti0YERGRgNO1YERERETioAJEREREPKcWjIiISMCFAngdKY2AiIiIiOdUgIiIiIjn1IIREREJOC1EJiIiIhIHFSAiIiLiObVgREREAi6ILZhyL0B+2FdQ3odIqma/vcPvCL8K19c6yu8ICcnbvtLvCAnbTZhmTZv6HSNhq1av9juCiHhALRgRERHxnFowIiIiARfEFoxGQERERMRzKkBERETEc2rBiIiIBFwIXQtGREREJCYVICIiIuI5tWBEREQCTmfBiIiIiMRBBYiIiIh4Ti0YERGRgAs5nQUjIiIiEpMKEBEREfGcWjAiIiIBp7NgREREROIQ2AIkJSWFl6e8waAXh/kdJS7du3dn5n/+w6zZs7nxxhv9jhNTEPKe/cCd3DZrAtdPfLFkW8ve3ciZOIJ7F00nq/UxPqaLLTMri5fGjmHKzOlMnv4uV1xztd+RYgrC50JEgiGwBchF11zJqhUr/I4Rl5SUFB4cOJCrr7qK3qefzjnnnMNRRx/td6wyBSXv4snvMvbGew7YtjF3FRPuGMD3C5b4lCp+RaEiHn/oYfr06s0lfc/nsiuv5Mijj/I7VpmC8rkQ+TUK4ZJ280rUAsTMTjazmpH7Vc3sATObYmaPmVktbyL+rwaZGXQ+rQdTXpvgV4SEtGvXjtWrVrFmzRoKCwuZMmUKZ/Tu7XesMgUl75oFX7Bn+44Dtm3+7nu2rFrrU6LEbNqwkWVfLgVg965drMzNpWFGps+pyhaUz4WIBEOsEZCRwO7I/aeAWsBjkW0vlWOuqG67/16efXQQ4XDYrwgJycjMZN369SWP169fT0Zmxf1FE7S8h4JGjQ+n1XGtWbJokd9RyqTPhYgkU6yzYFKcc0WR+9nOuQ6R+3PMrMyflGaWA+QAtKjXkIyatX950ojOPXuwdfMWvvlyKe1PPilp+y1PVso2V4EXjQla3qCrVq0aTw0byiMPDmRXQYHfccqkz4VIxeVl6yRZYo2AfGlm10TuLzazbAAzOwYoLOtNzrnhzrls51x2MosPgDYdO9C1V09e/2AmDz49mI6dTmHA4EFJPUay5eXl0Sgrq+RxVlYWG/LzfUwUXdDyBllaWhpPDhvK1DcnMeOdd/2OE5U+FyKSTLEKkOuA7ma2AmgNzDWzlcALkec8N2zQYPp26U6/br3of+sdLJg7jwfuuMuPKHFbvHgxzZo3p3GTJlSqVIk+ffowffp0v2OVKWh5g2zg44+xMjeXUS+O8DtKTPpciFRcoSTevBK1BeOc2w5cbWY1gBaR1691zum/PQkIhUL079+f0aNHk5qayrhx41i+fLnfscoUlLx9H7uPptltqVq7FrdMH8sHz41i7/YdnHHPLVSrU4uLnv07+V/nMvbGu/2OWqoO2dmc2+98vln2NRPfmgbAk4MG8cH7s/wNVoagfC5EpHyZ2UjgbGCDc+74yLa6wGtAM2AVcJFzbmvU/ZR3D7dzi2MD1ZhaF9rrd4RfhetrVdzTTUvz7+0r/Y6QsN0EY5L2z61avdrvCCLJUNq0qXLz1zanJe137UNL3o+a3cy6AQXA6P0KkMeBLc65R83sbqCOc+4v0fajpdhFREQCzsur4TrnPjCzZj/bfC7QI3J/FDALiFqABHYhMhEREUk+M8sxs/n73XLieFuGc249QOTPhrHeoBEQERERKeGcGw4ML+/jqAAREREJuAqwDki+mWU559abWRawIdYb1IIRERGRX2oycFXk/lXApFhvUAEiIiIicTOzV4G5wLFmttbMrgUeBXqb2XKgd+RxVGrBiIiIBJyXLRjn3KVlPNUrkf1oBEREREQ8pwJEREREPKcWjIiISMCFA3hlao2AiIiIiOdUgIiIiIjn1IIREREJuAqwEFnCyr0A0dVly9+dtY/xO0LC3t4Tc5G8CqV2amW/IyQsMyXV7wgJW1m4i2ZNm/odIyG6eq/IwVELRkRERDynFoyIiEjABbEFoxEQERER8ZwKEBEREfGcWjAiIiIBF9JCZCIiIiKxqQARERERz6kFIyIiEnA6C0ZEREQkDipARERExHNqwYiIiARcWGfBiIiIiMSmAkREREQ8pxaMiIhIwOksGBEREZE4BHIEpHv37vQfMIDU1FReGzuWoUOH+h0ppiBk7vnAn2jW7WT2bNnGq/1yAOh8+/U0734KocIitq9dx8z+/2Dfzl0+Jy1deo0a3PrwAJoecxQ4x5P3DODrRUv8jhVVSkoKIye9zsb8fO667g9+x4lpzKzp7N61i3AoTChUxI3nXeR3pJiC8N0T+TUKXAGSkpLCgwMHcsXll5OXl8fkyZOZPmMGucuX+x2tTEHJ/PWk9/ji1Umc/vCfS7atmfc5c58egQuF6XTbdXS89lLmPvmijynLlvPXP7Pgw4945NY7SauUxmFVqvodKaaLrrmSVStWkF69ut9R4nbHFVezY+s2v2PEJSjfPZFf6pBrwZjZrWbWxKsw8WjXrh2rV61izZo1FBYWMmXKFM7o3dvvWFEFJfO6z0H+nZMAABtySURBVL9g746dB2xbM3cBLhQGIH/JMqo3rO9HtJiqpqdzfHZH3hv/BgBFhUXs2rkzxrv81SAzg86n9WDKaxP8jnLICsp3T+TXKNYckIHAJ2b2oZn9PzNr4EWoaDIyM1m3fn3J4/Xr15ORmeljotiCmLk0rfr+htUffeZ3jFJlHdGY7Vu3cvujD/L0m69x68MDOKxqxR4Bue3+e3n20UGEw2G/o8TNOcegl19k2JvjOeviC/2OE9Oh8t0TORTFKkBWAo0pLkQ6Al+Z2TtmdpWZ1SjrTWaWY2bzzWz+zoKCJMYFK2Wbq+ALsAQx8891vO4ywqEQ306b6XeUUqWkpnJU65a8NWY8t/a9mL2793Bhzu/9jlWmzj17sHXzFr75cqnfURJy68WXc8O5F3D372+g7xWX0ubEjn5HiupQ+O6JxCPsXNJuXolVgDjnXNg5955z7lqgEfAccCbFxUlZbxrunMt2zmXXSHJvOy8vj0ZZWSWPs7Ky2JCfn9RjJFsQM++vZZ/eNO92MtPvedTvKGXanJfPprx8vlnyBQAfvTudo45r6XOqsrXp2IGuvXry+gczefDpwXTsdAoDBg/yO1ZMmzdsBGDbli3MmT6Tlm3a+JwouqB/90QOZbEKkAP+A+GcK3TOTXbOXQocUX6xyrZ48WKaNW9O4yZNqFSpEn369GH69Ol+RIlbEDP/5IjO2XS45mKm/rE/RXt/9DtOmbZu2szGvHwOb94UgLadTub73DJrZN8NGzSYvl26069bL/rfegcL5s7jgTvu8jtWVFWqVqVqerWS+9ldO/NdBZ/MGeTvnsihLtZZMBeX9YRzbk+Ss8QlFArRv39/Ro8eTWpqKuPGjWN5Bf8hGJTMZzx6L4dnt6FK7Vpc/d4YPhk6mo6/v4TUypU4d9hjAOR/sYxZDz3lc9LSPT/wUe76xyOkVapE3tq1PHl3f78jHVLq1K/Hg889DUBqWhozJ0/jsw/m+JwquqB890R+qSCeBWPl3Q9t1rRp8P5VAubO2sf4HSFhb+/Z4HeEhGwt2ut3hIRVSUn1O0LCVhZWzDVmolm1erXfEaRiKm0KUrm5pFV20n7Xjl0235PsWglVREREPBe4hchERETkQKEAnt2lERARERHxnAoQERER8ZxaMCIiIgEXDuBZMBoBEREREc+pABERERHPqQUjIiIScDoLRkRERCQOKkBERETEc2rBiIiIBFxYLRgRERGR2FSAiIiIiOfKvQXz5V9PK+9DJNXxD73vd4SE3TzmD35HSNiYPvf6HSEhj2c19DtCwnLWrvU7QsIapVbxO0JCPp7yd1j6ut8xEndcP78TSJKFtBCZiIiISGwqQERERMRzOgtGREQk4MIu7HeEhGkERERERDynAkREREQ8pxaMiIhIwIV1FoyIiIhIbCpARERExHNqwYiIiARcKIDXglEBIiIiInEzs1XATiAEFDnnsg9mPypAREREJFGnOec2/ZIdqAAREREJOJ0FIyIiIoFmZjlmNn+/W87PXuKA98xsQSnPxU0jICIiIlLCOTccGB7lJV2cc+vMrCEw3cy+ds59kOhxVICIiIgEXNjDs2Ccc+sif24wszeAk4BDswDJ2/Ej/d/OZfOuQlIMzmuTwWUds3j+ozW88UU+dapWAuCmU4+ga4s6PqctXffu3ek/YACpqam8NnYsQ4cO9TvSAX7cV8jlf32BfYVFhMJhftPpeG695HQuu+95du3ZB8Dm7QW0Oboxz939O5/Tli0lJYWRk15nY34+d133B7/j/I+j77mbOl06U7h1Kwt/d1XJ9qwL+pHV73xcKMTWj+ey6rmK9fkAyMzK4pEhT1C/QQNcOMy4Ma/yyksv+x0rLhX5c1HWd29N/hbuGDyW7QV7aN28EY//8UIqVwrEj2w5hJlZOpDinNsZuX8G8ODB7CsQn+bUFOP2Hk1plVGdXftCXPGvJZzStBYAl3VsxJUnNvI5YXQpKSk8OHAgV1x+OXl5eUyePJnpM2aQu3y539FKVK6UxqgHriW96mEUFoW47L7n6db+GMY8fEPJa255/N/0OrGVjylju+iaK1m1YgXp1av7HaVU+W+9zbrXJ3LM/feVbKvVoT31unZl4ZVX4woLqVS7to8Jy1YUKuLxhx5m2ZdLqZaezoSpU5g7Zw4rluf6HS2mivy5KOu799KUOVzdpwtndW1L/2FvMmHmfC478xS/44pkAG+YGRTXEGOcc+8czI6iTkI1s8pmdqWZnR55fJmZPWNmN5lZpYM54MFoUL0yrTKKf3CkV06led2qbCjY59Xhf7F27dqxetUq1qxZQ2FhIVOmTOGM3r39jnUAMyO96mEAFIVCFBWFiXzAACjY8yPzvljB6Se39itiTA0yM+h8Wg+mvDbB7yhl2rF4MUU7dhywLbNvX9a88gqusBCAwm3b/IgW06YNG1n25VIAdu/axcrcXBpmZPqcKraK/rko67s374uV/KbT8QCcd1oHZn66zM+YUsGFk3iLxjm30jnXNnI7zjn38MFmjjUC8lLkNdXM7CqgOjAR6EVxz+eqKO8tF+u27+XrDbs4Pqs6i3/YybiFeUxbupHWmenc3qMZNatUvEGdjMxM1q1fX/J4/fr1tGvf3sdEpQuFwpx/17N8n7eZy848hbbHNCl5bsa8pXQ64UiqV6viY8Lobrv/Xp59dBDV0tP9jpKQqkc0oVbbtjTLySG8bx/fPfMsBV9/7XesqBo1PpxWx7VmyaJFfkeJKQifi59/95pk1qVmehXSUlMByKxXk/zN231OKZJcsU7DPcE5dzFwHsV9ngucc/8CrgHK/A26/yk8Iz/4Nmlhd+8Lcdfkb7nztGZUPyyNC9plMOm69rx6VRvqp1dmyKxVSTtWMlkp21wFXDY3NTWFSYNvYfYLf2FJ7hq+XZ1X8tzUOUs469S2PqaLrnPPHmzdvIVvIv9DDxJLTSWtRg0W59zAd88+R8uBD/gdKapq1arx1LChPPLgQHYVFPgdJ6qgfC5+/t1buXbD/7xm/xFJkUNBrAIkxcwqAzWAakCtyPbDgDJbMM654c65bOdc9u+7HZOUoIWhMHdN/ob/a1WfnsfUA6BeemVSU4wUM85r05Cl6yvmD8O8vDwaZWWVPM7KymJDfr6PiaKrmV6Vk49rwYcLi+eobN25my+Wr6FHx2N9Tla2Nh070LVXT17/YCYPPj2Yjp1OYcDgQX7Hisu+DRvZPHs2AAXLluGcI62CzgNJS0vjyWFDmfrmJGa8867fcWIK2ufip+/eom/XsGPXXopCIQDyNu+gYd2aPqeTiizsXNJuXolVgIwAvgYWAfcB483sBeAzYGw5ZyvhnGPguytoXrcqV2T/d8Lpxv3mgby/fAtH1q/mVaSELF68mGbNm9O4SRMqVapEnz59mD59ut+xDrBlewE7du0BYO+PhXy8JJcWjRsA8M7HX9AjuyWHVfZs2k/Chg0aTN8u3enXrRf9b72DBXPn8cAdd/kdKy6bP/yQWh07AlClSRNS0tIoqqDzQAY+/hgrc3MZ9eIIv6PEJQifi9K+e0c2bsjJx7fg3blfAvDG+5/Ts4JPABdJVNQJE865IWb2WuT+OjMbDZwOvOCc+9SLgACLftjJtK82cVT9alw6ajFQfMrtu19v4psNuzCMRrUO497eLbyKlJBQKET//v0ZPXo0qampjBs3juUV6AwYgA1bd3L3PycQCjtcOMyZXU7gtOyWALw1ZwnXn9fd54SHhmP/NoBa7duTVrsWJ77xOt+PGEn+1Gkcfe89tP/XKFxhEd8+9He/Y5aqQ3Y25/Y7n2+Wfc3Et6YB8OSgQXzw/ix/gwVcWd+9oxo35PbBY3lyzHRaNW/Ehacf1PW+RCosK++5CAUvXF3xJjtEcfxD7/sdIWGr3hrsd4SEde5zr98REvJ4VkO/IyQsZ+1avyMkrHZqZb8jJOTjKRWzWIzpuH5+J/g18HTSTucWxybtd+3HK7/xJLuuBSMiIiKeUwEiIiIinqt4i2aIiIhIQrw8eyVZNAIiIiIinlMBIiIiIp5TC0ZERCTgwqgFIyIiIhKTChARERHxnFowIiIiAacWjIiIiEgcVICIiIiI59SCERERCbhw8DowGgERERER75X7CEj1zn3K+xBJ1Sv9a78jJGzNoEf8jpCw95++zO8ICTn2ppF+R0jYOTWa+B0hYU+/drvfERIyJ+dJvyMkbFFBFWCY3zEScvPi6X5HkHKgFoyIiEjA6SwYERERkTioABERERHPqQUjIiIScGrBiIiIiMRBBYiIiIh4Ti0YERGRgHPB68BoBERERES8pwJEREREPKcWjIiISMDpLBgRERGROKgAEREREc+pBSMiIhJwwWvAaAREREREfKACRERERDwXiBbMj/sKufyvL7CvsIhQOMxvOh3PrZeczitvzWXU1I/4Pm8Lc1++j7o10/2OWqbeV17CqRecA86x9tsVjLzvIYr27fM7VonUuhnUvf4BUmvVAxemYNYbFEwfS0p6Terd+Aip9bMIbVrPpufuxu3e6XdcAPK27uK+V+eyaedeUszod8qRXNGtJXeNnsOqjTsA2LmnkBpVKzH+T7/1OW3punfvTv8BA0hNTeW1sWMZOnSo35Gi6n7FhXTq1wczY+6Eycx6Zbzfkf7Hyh82cvsTY0ser8nfwq2XnE7fHu25/Ymx/LBxK4c3qMOTd15KrepVfUz6X0ffczd1unSmcOtWFv7uqpLtWRf0I6vf+bhQiK0fz2XVcxXj89HzgT/RrNvJ7NmyjVf75QDQ+fbrad79FEKFRWxfu46Z/f/Bvp27fE766xHEs2ACUYBUrpTGqAeuJb3qYRQWhbjsvufp1v4YOrRsSo/sllx5/wt+R4yqdsMG9LriIu7vcymFP/7IHwY/xMm/7c1Hb07zO1oJFypi29ghFK7+BqtSjYy//Yu9Sz8hvWsf9i77lJ3TRlHjrKuoedbVbB//T7/jApCamsKfzulA68Z12bW3kEuGvEOnY7IYdGXXktf8Y/LnVK9SyceUZUtJSeHBgQO54vLLycvLY/LkyUyfMYPc5cv9jlaqrKOa06lfH5649HpChUXcOOwJln4wl43fr/U72gFaHN6ASYNvASAUCtPt+kfpfXJrhr8xm05tjiTn/O4Mnzib4RNnc9eVZ/qctlj+W2+z7vWJHHP/fSXbanVoT72uXVl45dW4wkIq1a7tY8IDfT3pPb54dRKnP/znkm1r5n3O3KdH4EJhOt12HR2vvZS5T77oY0qp6GK2YMzsSDO708yeMrMnzOwPZlbLi3D7ZSC96mEAFIVCFBWFMTNat2hE44Z1vIxy0FJTU6lc5TBSUlOpXKUK2zZs9DvSAcLbN1O4+hsA3N7dFK1bRWqdhlRt351dc6YCsGvOVKp26OFjygM1qFmV1o3rApBepRLNM2qyYfvukuedc7y76Hv+r31TvyJG1a5dO1avWsWaNWsoLCxkypQpnNG7t9+xypTRohmrlyylcO+PhEMhcucvpE2vbn7HimruFytoklGXwxvWYeany+jboz0AfXu0Z8anX/mc7r92LF5M0Y4dB2zL7NuXNa+8gissBKBw2zY/opVq3edfsHfHgSOha+YuwIXCAOQvWUb1hvX9iPar5ZJ480rUAsTMbgWGAVWAE4GqQBNgrpn1KPd0+wmFwpx7xz/pfM3f6dz2KNoe08TLw/8i2zZs5N2X/s3jM99k8Oyp7CnYxdKPP/U7VplS62dRqemx7FvxJam16hLevhkoLlJSa1bMgu+HLQV8/cNWTmj63x96C1ZupF6NKjRtUNPHZGXLyMxk3fr1JY/Xr19PRmamj4miW5+7kiM7tqNarZpUqnIYrU/tRO3Mhn7HimranCWcfWpbADZvK6Bh3eLPQsO6NdmyvcDPaDFVPaIJtdq2pe3w5znhmX9SvWVLvyPFrVXf37D6o8/8jiEVXKwRkOuBM51zDwGnA62dc/cBZwJDynqTmeWY2Xwzmz98/PSkBE1NTWHS4FuY/cJfWJK7hm9X5yVlv16oVrMG7Xp24y+9z+dPPc7msKpVOKVPxRj6/Tk7rCr1b36cbWOewO0NRv9294+F3DHqQ/58bscD2i1vL1xVYUc/AKyUba4CX1Eqf+VqZox8hZteGMKNw57gh29zCYdCfscq077CIv7z2TLO7Hy831EOiqWmklajBotzbuC7Z5+j5cAH/I4Ul47XXUY4FOLbaTP9jiIVXDxnwfw0T+QwoAaAc+57oMzGunNuuHMu2zmXnXNhcoeUa6ZX5eTjWvDhworZJy9N604nsumHdRRs3UaoKMSC6bM4qt0Jfsf6X6mp1Lv5cXbNfYc9C94HILR9Cym16gGQUqseoR1b/Uz4PwpDYe54+UPO6tCM09v8d1SsKBRm5hdr+U27iluA5OXl0Sgrq+RxVlYWG/LzfUwU27yJ0xh00bU8ffXN7N6+g42rK9b8j/19sPBbjmvRiPq1awBQr3Z1NmwpbnNs2LKDurWq+xkvpn0bNrJ59mwACpYtwzlHWgWaB1Kaln1607zbyUy/51G/o/zqHHItGOBF4DMzGw7MBZ4BMLMGwJZyzlZiy/YCduzaA8DeHwv5eEkuLRo38Orwv9jm9fm0aHs8lasUz2NpdUo261au8jdUKer+vj9F67+j4N1/l2zbs2g26V3PBiC969nsWTjbr3j/wznHgNfm0TyjFld2b3XAc/OW59G8YU0ya1fzKV1sixcvplnz5jRu0oRKlSrRp08fpk9Pzohhealet/gXYJ3MDNr26s6Ct2f4nKhs0z5czFld25Y87nliK96ctRCAN2ctpNdJrcp6a4Ww+cMPqdWxIwBVmjQhJS2Nogo0D+TnjuicTYdrLmbqH/tTtPdHv+NIAEQ9C8Y595SZzQBaAYOdc19Htm8EPJt9tmHrTu7+5wRCYYcLhzmzywmclt2S0dM+5sU3PmDTtgLOuf1punc4lodvOt+rWHH7bslSFrz3H/pPGEU4FOL7Zd/ywbg3/Y51gMpHtyW9y1nsW7OcjAeLC5DtE55j59RR1LvpEdJPPZfQljw2P3u3z0n/a+F3G5m6YBVHZ9XmwifeAuDW37bl1FaH887C1RW6/QIQCoXo378/o0ePJjU1lXHjxrG8gp4B85NrhzxMeu2ahIpCjH94MHt2VIxTsn9uz4/7+HhxLg/+4bySbTnnd+e2f4xhwsz5ZNWvxVN3XuZjwgMd+7cB1GrfnrTatTjxjdf5fsRI8qdO4+h776H9v0bhCov49qG/+x2zxBmP3svh2W2oUrsWV783hk+Gjqbj7y8htXIlzh32GAD5Xyxj1kNP+ZxUKjIr957z0tcrblO7FNdeOMjvCAn720lFfkdIWMML+vgdISHH3jTS7wgJO6dGcCZq/+Tp1273O0JC5uQ86XeEhC0qqOJ3hITdvLhijwyWobRpXuXmqKZNk/a7Nnf1ak+yayVUERER8ZwKEBEREfFcIFZCFRERkbIFaq5DhEZARERExHMqQERERMRzasGIiIgEnFowIiIiInFQASIiIiKeUwtGREQk4NSCEREREYmDChARERHxnAoQERGRgHNJvMViZmea2TdmlmtmB32FUhUgIiIiEhczSwWeBf4PaA1camatD2ZfKkBEREQkXicBuc65lc65fcBY4NyD2ZE5F8S5s8XMLMc5N9zvHPEKWl4IXuag5QVl9kLQ8oIyeyFoeb1iZjlAzn6bhv/072RmFwBnOueuizz+HXCyc+7mRI8T9BGQnNgvqVCClheClzloeUGZvRC0vKDMXghaXk8454Y757L3u+1fpFlpbzmY4wS9ABERERHvrAWa7Pe4MbDuYHakAkRERETi9RlwtJk1N7PKwCXA5IPZUdBXQg1a7y5oeSF4mYOWF5TZC0HLC8rshaDl9Z1zrsjMbgbeBVKBkc65pQezr0BPQhUREZFgUgtGREREPKcCRERERDwXyAIkWcvAesXMRprZBjP70u8s8TCzJmb2vpktM7OlZvZHvzPFYmZVzOxTM1scyfyA35niYWapZrbQzKb6nSUeZrbKzL4ws0VmNt/vPPEws9pmNsHMvo58pjv5nSkaMzs28u/7022Hmd3md65ozOz2yPfuSzN71cyq+J0pFjP7YyTv0or+73uoCtwckMgysN8CvSk+Hegz4FLn3Fe+BovCzLoBBcBo59zxfueJxcyygCzn3OdmVgNYAPSt4P/GBqQ75wrMrBIwB/ijc26ez9GiMrM7gGygpnPubL/zxGJmq4Bs59wmv7PEy8xGAR86516MzNqv5pzb5neueER+3v1A8UJPq/3OUxozO5zi71tr59weMxsHvOWce9nfZGUzs+MpXsHzJGAf8A5wo3Nuua/BfmWCOAKStGVgveKc+wDY4neOeDnn1jvnPo/c3wksAw73N1V0rlhB5GGlyK1CV9dm1hg4C3jR7yyHKjOrCXQDRgA45/YFpfiI6AWsqKjFx37SgKpmlgZU4yDXhfBQK2Cec263c64ImA2c53OmX50gFiCHA2v2e7yWCv7LMcjMrBnQHvjE3ySxRdoZi4ANwHTnXEXP/CTwZyDsd5AEOOA9M1sQWa65omsBbAReirS6XjSzdL9DJeAS4FW/Q0TjnPsB+AfwPbAe2O6ce8/fVDF9CXQzs3pmVg34LQcuriUeCGIBkrRlYCU6M6sOvA7c5pzb4XeeWJxzIedcO4pX5jspMsxaIZnZ2cAG59wCv7MkqItzrgPFV8K8KdJerMjSgA7AUOdce2AXUOHnjQFE2kXnAOP9zhKNmdWheBS6OdAISDezK/xNFZ1zbhnwGDCd4vbLYqDI11C/QkEsQJK2DKyULTKP4nXg3865iX7nSURkiH0WcKbPUaLpApwTmVMxFuhpZq/4Gyk259y6yJ8bgDcobolWZGuBtfuNhk2guCAJgv8DPnfO5fsdJIbTge+ccxudc4XARKCzz5lics6NcM51cM51o7hFrvkfHgtiAZK0ZWCldJEJnSOAZc65wX7niYeZNTCz2pH7VSn+ofi1v6nK5py7xznX2DnXjOLP8H+ccxX6f41mlh6ZlEykjXEGxUPZFZZzLg9YY2bHRjb1AirsZOqfuZQK3n6J+B44xcyqRX529KJ43liFZmYNI38eAZxPMP6tDymBW4o9mcvAesXMXgV6APXNbC0wwDk3wt9UUXUBfgd8EZlTAXCvc+4tHzPFkgWMipw1kAKMc84F4tTWAMkA3ij+HUMaMMY5946/keJyC/DvyH9YVgLX+Jwnpsi8hN7ADX5nicU594mZTQA+p7iNsZBgLHH+upnVAwqBm5xzW/0O9GsTuNNwRUREJPiC2IIRERGRgFMBIiIiIp5TASIiIiKeUwEiIiIinlMBIiIiIp5TASIiIiKeUwEiIiIinvv/AZQCeujkjhEAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model_assess(gs_NB.best_estimator_,\"GaussianNB GridSearchCV\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "F3TWMGsuDTAF"
   },
   "source": [
    "## ROC for Grid Search NB"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 295
    },
    "id": "vd9504La24aW",
    "outputId": "06110b47-4c61-499c-b3c2-fff507b39726"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ROC(gs_NB)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "H33NzRdxDbji"
   },
   "source": [
    "# SGD Classifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "cwMHhxgqDvRD"
   },
   "source": [
    "## Simple SGD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "rpg2UWn8DyfK",
    "outputId": "cf62bd88-225f-489b-890f-f2c5af911750"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Default :\n",
      "Accuracy score: 0.6041041041041041\n",
      "Overall Precision: 0.5972550877132848\n",
      "Overall Recall:0.6103587617280029\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.51      0.74      0.60       191\n",
      "           1       0.90      0.92      0.91       214\n",
      "           2       0.46      0.59      0.52       193\n",
      "           3       0.51      0.49      0.50       202\n",
      "           4       0.53      0.55      0.54       201\n",
      "           5       0.67      0.62      0.65       202\n",
      "           6       0.73      0.87      0.80       187\n",
      "           7       0.85      0.67      0.75       177\n",
      "           8       0.49      0.56      0.52       211\n",
      "           9       0.31      0.09      0.14       220\n",
      "\n",
      "    accuracy                           0.60      1998\n",
      "   macro avg       0.60      0.61      0.59      1998\n",
      "weighted avg       0.59      0.60      0.59      1998\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import SGDClassifier\n",
    "sgd=SGDClassifier()\n",
    "sgd.fit(X_train,y_train)\n",
    "model_assess(sgd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "id": "VRbOMuWU3Cfv"
   },
   "outputs": [],
   "source": [
    "loss = ['hinge', 'log', 'modified_huber', 'squared_hinge',  \n",
    "'perceptron'] \n",
    "penalty = ['l1', 'l2', 'elasticnet'] \n",
    "alpha = [0.001, 0.01, 0.1, 1, 10, 100] \n",
    "learning_rate = ['constant', 'optimal', 'invscaling', 'adaptive'] \n",
    "\n",
    "param_distributions = dict(loss=loss, \n",
    "penalty=penalty, \n",
    "alpha=alpha, \n",
    "learning_rate=learning_rate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Rt-Dr-iy3kbP",
    "outputId": "7bf47f4d-4edf-4fb8-9d27-6b9b752e265b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 360 candidates, totalling 1800 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n",
      "/home/ashita/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/_stochastic_gradient.py:700: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  ConvergenceWarning,\n"
     ]
    }
   ],
   "source": [
    "model_sgd=GridSearchCV(estimator=SGDClassifier(), \n",
    "                      param_grid=param_distributions, \n",
    "                      scoring='accuracy', \n",
    "                      verbose=1)\n",
    "model_sgd.fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "0KqxKUsO9imR",
    "outputId": "0543f570-b269-48ed-e349-ee576ff14b4e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SGD classifier :\n",
      "Accuracy score: 0.6441441441441441\n",
      "Overall Precision: 0.6386137102787109\n",
      "Overall Recall:0.6421140902032518\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.56      0.64      0.60       195\n",
      "           1       0.89      0.93      0.91       199\n",
      "           2       0.52      0.49      0.51       194\n",
      "           3       0.52      0.53      0.53       203\n",
      "           4       0.62      0.59      0.61       201\n",
      "           5       0.65      0.69      0.67       186\n",
      "           6       0.82      0.82      0.82       208\n",
      "           7       0.73      0.75      0.74       202\n",
      "           8       0.58      0.61      0.60       223\n",
      "           9       0.47      0.36      0.41       187\n",
      "\n",
      "    accuracy                           0.64      1998\n",
      "   macro avg       0.64      0.64      0.64      1998\n",
      "weighted avg       0.64      0.64      0.64      1998\n",
      "\n"
     ]
    }
   ],
   "source": [
    "model_assess(model_sgd.best_estimator_,\"SGD classifier\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "HCy7iIkUFI6t"
   },
   "source": [
    "## ROC for SGD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 295
    },
    "id": "QpDlbGNM9mYg",
    "outputId": "4f606ff1-b9f0-447f-e144-941dd1b729a5"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ROC(model_sgd.best_estimator_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Best Model :\n",
    "- KNN with GridSearch Accuracy: 90%\n",
    "{'metric': 'manhattan', 'n_neighbors': 1, 'weights': 'uniform'}"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "NB SGD KNN.ipynb",
   "provenance": []
  },
  "interpreter": {
   "hash": "691a1f0c4a2cc4687a0c7a533259563740bfdd0381bbb1595843813bdc6b6e58"
  },
  "kernelspec": {
   "display_name": "Python 3.7.6 64-bit ('base': conda)",
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}