sentiment-analysis-svm-nb-bagging-boosting-rf-lstm.ipynb

{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.7.10","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\">Twitter Sentiment Analysis With SVM, Naive Bayes, Bagging, Random Forest, And LSTM  </h1>\n\n</div>\n<p><img src = \"https://www.vshsolutions.com/wp-content/uploads/2019/02/airlines.jpeg\"  ></p>\n","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2021-10-08T12:58:11.812144Z","iopub.execute_input":"2021-10-08T12:58:11.812902Z","iopub.status.idle":"2021-10-08T12:58:11.829075Z","shell.execute_reply.started":"2021-10-08T12:58:11.812847Z","shell.execute_reply":"2021-10-08T12:58:11.828148Z"}}},{"cell_type":"markdown","source":"<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\"> Importing Libraries </h1>\n</div>\n","metadata":{}},{"cell_type":"code","source":"import pandas as pd\nimport seaborn as sns\nimport re, nltk\nnltk.download('punkt')\nimport matplotlib.pyplot as plt \nfrom sklearn.model_selection import train_test_split, StratifiedKFold, cross_val_score\nfrom sklearn import model_selection, naive_bayes, svm\nfrom sklearn.metrics import classification_report,confusion_matrix\nfrom sklearn.metrics import roc_auc_score\nfrom sklearn.metrics import recall_score\nfrom sklearn.metrics import f1_score\nfrom sklearn.metrics import confusion_matrix, accuracy_score\nfrom sklearn.model_selection import GridSearchCV\nfrom sklearn.metrics import precision_recall_curve\nfrom sklearn.metrics import f1_score\nfrom sklearn.metrics import auc\nfrom matplotlib import pyplot\nfrom sklearn.metrics import roc_curve\nfrom sklearn.metrics import roc_auc_score, accuracy_score\nimport string\nfrom nltk.corpus import stopwords\nnltk.download('stopwords')\nfrom sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer\nfrom sklearn.naive_bayes import MultinomialNB, GaussianNB\nfrom sklearn.metrics import f1_score\nplt.style.use('fivethirtyeight')\nplt.style.use('dark_background')\nfrom sklearn.ensemble import AdaBoostClassifier\nfrom sklearn.model_selection import cross_val_score\nimport numpy as np\nfrom sklearn.ensemble import BaggingClassifier\nfrom sklearn.ensemble import RandomForestClassifier\nfrom lime import lime_tabular\nfrom tensorflow.keras.layers import Embedding\nfrom tensorflow.keras.preprocessing.sequence import pad_sequences\nfrom tensorflow.keras.models import Sequential\nfrom tensorflow.keras.preprocessing.text import one_hot\nfrom tensorflow.keras.layers import LSTM, Bidirectional\nfrom tensorflow.keras.layers import Dense, Dropout\n\n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:19:23.688808Z","iopub.execute_input":"2022-01-06T07:19:23.689094Z","iopub.status.idle":"2022-01-06T07:19:29.655528Z","shell.execute_reply.started":"2022-01-06T07:19:23.689021Z","shell.execute_reply":"2022-01-06T07:19:29.654784Z"},"trusted":true},"execution_count":1,"outputs":[{"name":"stdout","text":"[nltk_data] Downloading package punkt to /usr/share/nltk_data...\n[nltk_data]   Package punkt is already up-to-date!\n[nltk_data] Downloading package stopwords to /usr/share/nltk_data...\n[nltk_data]   Unzipping corpora/stopwords.zip.\n","output_type":"stream"},{"name":"stderr","text":"2022-01-06 07:19:25.890915: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0\n","output_type":"stream"}]},{"cell_type":"code","source":"df =pd.read_csv(r'../input/twitter-airline-sentiment/Tweets.csv')\ndf.head()","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:19:29.658775Z","iopub.execute_input":"2022-01-06T07:19:29.658972Z","iopub.status.idle":"2022-01-06T07:19:29.800728Z","shell.execute_reply.started":"2022-01-06T07:19:29.658949Z","shell.execute_reply":"2022-01-06T07:19:29.800061Z"},"trusted":true},"execution_count":2,"outputs":[{"execution_count":2,"output_type":"execute_result","data":{"text/plain":"             tweet_id airline_sentiment  airline_sentiment_confidence  \\\n0  570306133677760513           neutral                        1.0000   \n1  570301130888122368          positive                        0.3486   \n2  570301083672813571           neutral                        0.6837   \n3  570301031407624196          negative                        1.0000   \n4  570300817074462722          negative                        1.0000   \n\n  negativereason  negativereason_confidence         airline  \\\n0            NaN                        NaN  Virgin America   \n1            NaN                     0.0000  Virgin America   \n2            NaN                        NaN  Virgin America   \n3     Bad Flight                     0.7033  Virgin America   \n4     Can't Tell                     1.0000  Virgin America   \n\n  airline_sentiment_gold        name negativereason_gold  retweet_count  \\\n0                    NaN     cairdin                 NaN              0   \n1                    NaN    jnardino                 NaN              0   \n2                    NaN  yvonnalynn                 NaN              0   \n3                    NaN    jnardino                 NaN              0   \n4                    NaN    jnardino                 NaN              0   \n\n                                                text tweet_coord  \\\n0                @VirginAmerica What @dhepburn said.         NaN   \n1  @VirginAmerica plus you've added commercials t...         NaN   \n2  @VirginAmerica I didn't today... Must mean I n...         NaN   \n3  @VirginAmerica it's really aggressive to blast...         NaN   \n4  @VirginAmerica and it's a really big bad thing...         NaN   \n\n               tweet_created tweet_location               user_timezone  \n0  2015-02-24 11:35:52 -0800            NaN  Eastern Time (US & Canada)  \n1  2015-02-24 11:15:59 -0800            NaN  Pacific Time (US & Canada)  \n2  2015-02-24 11:15:48 -0800      Lets Play  Central Time (US & Canada)  \n3  2015-02-24 11:15:36 -0800            NaN  Pacific Time (US & Canada)  \n4  2015-02-24 11:14:45 -0800            NaN  Pacific Time (US & Canada)  ","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>tweet_id</th>\n      <th>airline_sentiment</th>\n      <th>airline_sentiment_confidence</th>\n      <th>negativereason</th>\n      <th>negativereason_confidence</th>\n      <th>airline</th>\n      <th>airline_sentiment_gold</th>\n      <th>name</th>\n      <th>negativereason_gold</th>\n      <th>retweet_count</th>\n      <th>text</th>\n      <th>tweet_coord</th>\n      <th>tweet_created</th>\n      <th>tweet_location</th>\n      <th>user_timezone</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>570306133677760513</td>\n      <td>neutral</td>\n      <td>1.0000</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>cairdin</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica What @dhepburn said.</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:35:52 -0800</td>\n      <td>NaN</td>\n      <td>Eastern Time (US &amp; Canada)</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>570301130888122368</td>\n      <td>positive</td>\n      <td>0.3486</td>\n      <td>NaN</td>\n      <td>0.0000</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>jnardino</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica plus you've added commercials t...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:15:59 -0800</td>\n      <td>NaN</td>\n      <td>Pacific Time (US &amp; Canada)</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>570301083672813571</td>\n      <td>neutral</td>\n      <td>0.6837</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>yvonnalynn</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica I didn't today... Must mean I n...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:15:48 -0800</td>\n      <td>Lets Play</td>\n      <td>Central Time (US &amp; Canada)</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>570301031407624196</td>\n      <td>negative</td>\n      <td>1.0000</td>\n      <td>Bad Flight</td>\n      <td>0.7033</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>jnardino</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica it's really aggressive to blast...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:15:36 -0800</td>\n      <td>NaN</td>\n      <td>Pacific Time (US &amp; Canada)</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>570300817074462722</td>\n      <td>negative</td>\n      <td>1.0000</td>\n      <td>Can't Tell</td>\n      <td>1.0000</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>jnardino</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica and it's a really big bad thing...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:14:45 -0800</td>\n      <td>NaN</td>\n      <td>Pacific Time (US &amp; Canada)</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"markdown","source":"","metadata":{}},{"cell_type":"code","source":"# Unique values of sentiment\ndf['airline_sentiment'].unique()","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:19:29.801793Z","iopub.execute_input":"2022-01-06T07:19:29.802425Z","iopub.status.idle":"2022-01-06T07:19:29.812622Z","shell.execute_reply.started":"2022-01-06T07:19:29.802388Z","shell.execute_reply":"2022-01-06T07:19:29.811929Z"},"trusted":true},"execution_count":3,"outputs":[{"execution_count":3,"output_type":"execute_result","data":{"text/plain":"array(['neutral', 'positive', 'negative'], dtype=object)"},"metadata":{}}]},{"cell_type":"code","source":"# Unique values of sentiment plot\n\nax = sns.countplot(x=\"airline_sentiment\", data=df)","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:19:29.815972Z","iopub.execute_input":"2022-01-06T07:19:29.816619Z","iopub.status.idle":"2022-01-06T07:19:30.020444Z","shell.execute_reply.started":"2022-01-06T07:19:29.816591Z","shell.execute_reply":"2022-01-06T07:19:30.019659Z"},"trusted":true},"execution_count":4,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":"> Positive and neutral tweets are almost equal.\n\n> Negative tweets are more than double of neutral or positive sentiments.","metadata":{}},{"cell_type":"code","source":"# Unique values of airline\n\nplt.figure(figsize=(10,10))\nax = sns.countplot(x=\"airline\", data=df)","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:19:30.021605Z","iopub.execute_input":"2022-01-06T07:19:30.021988Z","iopub.status.idle":"2022-01-06T07:19:30.256143Z","shell.execute_reply.started":"2022-01-06T07:19:30.021954Z","shell.execute_reply":"2022-01-06T07:19:30.255496Z"},"trusted":true},"execution_count":5,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 720x720 with 1 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":"> United has the most number of flights.\n\n> Virgin America has the least.","metadata":{}},{"cell_type":"markdown","source":"<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\"> Utility Functions </h1>\n</div>\n","metadata":{}},{"cell_type":"code","source":"# I am tokenizing the tweet and also taking tokens from second index onwards as initital to gives airline name and '@' and lowering thm and later making it back a sentence\ndef clean_the_tweet(text):\n  tokens= nltk.word_tokenize(re.sub(\"[^a-zA-Z]\", \" \",text))\n  tokens = [token.lower() for token in tokens]\n  return ' '.join(tokens[2:])\n\n                 \n\ndef text_process(msg):\n  nopunc =[char for char in msg if char not in string.punctuation]\n  nopunc=''.join(nopunc)\n  return ' '.join([word for word in nopunc.split() if word.lower() not in stopwords.words('english')])\n\n\n \ndef check_scores(clf,X_train, X_test, y_train, y_test):\n\n  model=clf.fit(X_train, y_train)\n  predicted_class=model.predict(X_test)\n  predicted_class_train=model.predict(X_train)\n  test_probs = model.predict_proba(X_test)\n  test_probs = test_probs[:, 1]\n  yhat = model.predict(X_test)\n  lr_precision, lr_recall, _ = precision_recall_curve(y_test, test_probs)\n  lr_f1, lr_auc = f1_score(y_test, yhat), auc(lr_recall, lr_precision)\n\n\n  print('Train confusion matrix is: ',)\n  print(confusion_matrix(y_train, predicted_class_train))\n\n  print()\n  print('Test confusion matrix is: ')\n  print(confusion_matrix(y_test, predicted_class))\n  print()\n  print(classification_report(y_test,predicted_class)) \n  print() \n  train_accuracy = accuracy_score(y_train,predicted_class_train)\n  test_accuracy = accuracy_score(y_test,predicted_class)\n\n  print(\"Train accuracy score: \", train_accuracy)\n  print(\"Test accuracy score: \",test_accuracy )\n  print()\n  train_auc = roc_auc_score(y_train, clf.predict_proba(X_train)[:,1])\n  test_auc = roc_auc_score(y_test, clf.predict_proba(X_test)[:,1])\n\n  print(\"Train ROC-AUC score: \", train_auc)\n  print(\"Test ROC-AUC score: \", test_auc)\n  fig, (ax1, ax2) = plt.subplots(1, 2)\n\n  ax1.plot(lr_recall, lr_precision)\n  ax1.set(xlabel=\"Recall\", ylabel=\"Precision\")\n\n  plt.subplots_adjust(left=0.5,\n                    bottom=0.1, \n                    right=1.5, \n                    top=0.9, \n                    wspace=0.4, \n                    hspace=0.4)\n  print()\n  print('Are under Precision-Recall curve:', lr_f1)\n  \n  fpr, tpr, _ = roc_curve(y_test, test_probs)\n\n\n  ax2.plot(fpr, tpr)\n  ax2.set(xlabel='False Positive Rate', ylabel='True Positive Rate')\n\n  print(\"Area under ROC-AUC:\", lr_auc)\n  return train_accuracy, test_accuracy, train_auc, test_auc\n\n\n\ndef grid_search(model, parameters, X_train, Y_train):\n  #Doing a grid\n  grid = GridSearchCV(estimator=model,\n                       param_grid = parameters,\n                       cv = 2, verbose=2, scoring='roc_auc')\n  #Fitting the grid \n  grid.fit(X_train,Y_train)\n  print()\n  print()\n  # Best model found using grid search\n  optimal_model = grid.best_estimator_\n  print('Best parameters are: ')\n  print( grid.best_params_)\n\n  return optimal_model\n  \n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:19:30.257517Z","iopub.execute_input":"2022-01-06T07:19:30.257964Z","iopub.status.idle":"2022-01-06T07:19:30.274647Z","shell.execute_reply.started":"2022-01-06T07:19:30.257928Z","shell.execute_reply":"2022-01-06T07:19:30.273912Z"},"trusted":true},"execution_count":6,"outputs":[]},{"cell_type":"markdown","source":"<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\"> Text Preparation </h1>\n</div>\n","metadata":{}},{"cell_type":"code","source":"# removing neutral tweets\n\ndf = df[df['airline_sentiment']!='neutral']\ndf['cleaned_tweet'] = df['text'].apply(clean_the_tweet)\n\ndf.head()\ndf['airline_sentiment'] = df['airline_sentiment'].apply(lambda x: 1 if x =='positive' else 0)\ndf.head()","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:19:30.275754Z","iopub.execute_input":"2022-01-06T07:19:30.276071Z","iopub.status.idle":"2022-01-06T07:19:32.796339Z","shell.execute_reply.started":"2022-01-06T07:19:30.276036Z","shell.execute_reply":"2022-01-06T07:19:32.795646Z"},"trusted":true},"execution_count":7,"outputs":[{"execution_count":7,"output_type":"execute_result","data":{"text/plain":"             tweet_id  airline_sentiment  airline_sentiment_confidence  \\\n1  570301130888122368                  1                        0.3486   \n3  570301031407624196                  0                        1.0000   \n4  570300817074462722                  0                        1.0000   \n5  570300767074181121                  0                        1.0000   \n6  570300616901320704                  1                        0.6745   \n\n  negativereason  negativereason_confidence         airline  \\\n1            NaN                     0.0000  Virgin America   \n3     Bad Flight                     0.7033  Virgin America   \n4     Can't Tell                     1.0000  Virgin America   \n5     Can't Tell                     0.6842  Virgin America   \n6            NaN                     0.0000  Virgin America   \n\n  airline_sentiment_gold        name negativereason_gold  retweet_count  \\\n1                    NaN    jnardino                 NaN              0   \n3                    NaN    jnardino                 NaN              0   \n4                    NaN    jnardino                 NaN              0   \n5                    NaN    jnardino                 NaN              0   \n6                    NaN  cjmcginnis                 NaN              0   \n\n                                                text tweet_coord  \\\n1  @VirginAmerica plus you've added commercials t...         NaN   \n3  @VirginAmerica it's really aggressive to blast...         NaN   \n4  @VirginAmerica and it's a really big bad thing...         NaN   \n5  @VirginAmerica seriously would pay $30 a fligh...         NaN   \n6  @VirginAmerica yes, nearly every time I fly VX...         NaN   \n\n               tweet_created    tweet_location               user_timezone  \\\n1  2015-02-24 11:15:59 -0800               NaN  Pacific Time (US & Canada)   \n3  2015-02-24 11:15:36 -0800               NaN  Pacific Time (US & Canada)   \n4  2015-02-24 11:14:45 -0800               NaN  Pacific Time (US & Canada)   \n5  2015-02-24 11:14:33 -0800               NaN  Pacific Time (US & Canada)   \n6  2015-02-24 11:13:57 -0800  San Francisco CA  Pacific Time (US & Canada)   \n\n                                       cleaned_tweet  \n1   you ve added commercials to the experience tacky  \n3  s really aggressive to blast obnoxious enterta...  \n4               it s a really big bad thing about it  \n5  would pay a flight for seats that didn t have ...  \n6  nearly every time i fly vx this ear worm won t...  ","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>tweet_id</th>\n      <th>airline_sentiment</th>\n      <th>airline_sentiment_confidence</th>\n      <th>negativereason</th>\n      <th>negativereason_confidence</th>\n      <th>airline</th>\n      <th>airline_sentiment_gold</th>\n      <th>name</th>\n      <th>negativereason_gold</th>\n      <th>retweet_count</th>\n      <th>text</th>\n      <th>tweet_coord</th>\n      <th>tweet_created</th>\n      <th>tweet_location</th>\n      <th>user_timezone</th>\n      <th>cleaned_tweet</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>1</th>\n      <td>570301130888122368</td>\n      <td>1</td>\n      <td>0.3486</td>\n      <td>NaN</td>\n      <td>0.0000</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>jnardino</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica plus you've added commercials t...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:15:59 -0800</td>\n      <td>NaN</td>\n      <td>Pacific Time (US &amp; Canada)</td>\n      <td>you ve added commercials to the experience tacky</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>570301031407624196</td>\n      <td>0</td>\n      <td>1.0000</td>\n      <td>Bad Flight</td>\n      <td>0.7033</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>jnardino</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica it's really aggressive to blast...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:15:36 -0800</td>\n      <td>NaN</td>\n      <td>Pacific Time (US &amp; Canada)</td>\n      <td>s really aggressive to blast obnoxious enterta...</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>570300817074462722</td>\n      <td>0</td>\n      <td>1.0000</td>\n      <td>Can't Tell</td>\n      <td>1.0000</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>jnardino</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica and it's a really big bad thing...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:14:45 -0800</td>\n      <td>NaN</td>\n      <td>Pacific Time (US &amp; Canada)</td>\n      <td>it s a really big bad thing about it</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>570300767074181121</td>\n      <td>0</td>\n      <td>1.0000</td>\n      <td>Can't Tell</td>\n      <td>0.6842</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>jnardino</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica seriously would pay $30 a fligh...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:14:33 -0800</td>\n      <td>NaN</td>\n      <td>Pacific Time (US &amp; Canada)</td>\n      <td>would pay a flight for seats that didn t have ...</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>570300616901320704</td>\n      <td>1</td>\n      <td>0.6745</td>\n      <td>NaN</td>\n      <td>0.0000</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>cjmcginnis</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica yes, nearly every time I fly VX...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:13:57 -0800</td>\n      <td>San Francisco CA</td>\n      <td>Pacific Time (US &amp; Canada)</td>\n      <td>nearly every time i fly vx this ear worm won t...</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"# Cleaning the tweets, removing punctuation marks\ndf['cleaned_tweet'] = df['cleaned_tweet'].apply(text_process)\ndf.reset_index(drop=True, inplace = True)\ndf.head()","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:19:32.797466Z","iopub.execute_input":"2022-01-06T07:19:32.798315Z","iopub.status.idle":"2022-01-06T07:19:55.616686Z","shell.execute_reply.started":"2022-01-06T07:19:32.798278Z","shell.execute_reply":"2022-01-06T07:19:55.616021Z"},"trusted":true},"execution_count":8,"outputs":[{"execution_count":8,"output_type":"execute_result","data":{"text/plain":"             tweet_id  airline_sentiment  airline_sentiment_confidence  \\\n0  570301130888122368                  1                        0.3486   \n1  570301031407624196                  0                        1.0000   \n2  570300817074462722                  0                        1.0000   \n3  570300767074181121                  0                        1.0000   \n4  570300616901320704                  1                        0.6745   \n\n  negativereason  negativereason_confidence         airline  \\\n0            NaN                     0.0000  Virgin America   \n1     Bad Flight                     0.7033  Virgin America   \n2     Can't Tell                     1.0000  Virgin America   \n3     Can't Tell                     0.6842  Virgin America   \n4            NaN                     0.0000  Virgin America   \n\n  airline_sentiment_gold        name negativereason_gold  retweet_count  \\\n0                    NaN    jnardino                 NaN              0   \n1                    NaN    jnardino                 NaN              0   \n2                    NaN    jnardino                 NaN              0   \n3                    NaN    jnardino                 NaN              0   \n4                    NaN  cjmcginnis                 NaN              0   \n\n                                                text tweet_coord  \\\n0  @VirginAmerica plus you've added commercials t...         NaN   \n1  @VirginAmerica it's really aggressive to blast...         NaN   \n2  @VirginAmerica and it's a really big bad thing...         NaN   \n3  @VirginAmerica seriously would pay $30 a fligh...         NaN   \n4  @VirginAmerica yes, nearly every time I fly VX...         NaN   \n\n               tweet_created    tweet_location               user_timezone  \\\n0  2015-02-24 11:15:59 -0800               NaN  Pacific Time (US & Canada)   \n1  2015-02-24 11:15:36 -0800               NaN  Pacific Time (US & Canada)   \n2  2015-02-24 11:14:45 -0800               NaN  Pacific Time (US & Canada)   \n3  2015-02-24 11:14:33 -0800               NaN  Pacific Time (US & Canada)   \n4  2015-02-24 11:13:57 -0800  San Francisco CA  Pacific Time (US & Canada)   \n\n                                       cleaned_tweet  \n0                 added commercials experience tacky  \n1  really aggressive blast obnoxious entertainmen...  \n2                               really big bad thing  \n3  would pay flight seats playing really bad thin...  \n4          nearly every time fly vx ear worm go away  ","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>tweet_id</th>\n      <th>airline_sentiment</th>\n      <th>airline_sentiment_confidence</th>\n      <th>negativereason</th>\n      <th>negativereason_confidence</th>\n      <th>airline</th>\n      <th>airline_sentiment_gold</th>\n      <th>name</th>\n      <th>negativereason_gold</th>\n      <th>retweet_count</th>\n      <th>text</th>\n      <th>tweet_coord</th>\n      <th>tweet_created</th>\n      <th>tweet_location</th>\n      <th>user_timezone</th>\n      <th>cleaned_tweet</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>570301130888122368</td>\n      <td>1</td>\n      <td>0.3486</td>\n      <td>NaN</td>\n      <td>0.0000</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>jnardino</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica plus you've added commercials t...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:15:59 -0800</td>\n      <td>NaN</td>\n      <td>Pacific Time (US &amp; Canada)</td>\n      <td>added commercials experience tacky</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>570301031407624196</td>\n      <td>0</td>\n      <td>1.0000</td>\n      <td>Bad Flight</td>\n      <td>0.7033</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>jnardino</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica it's really aggressive to blast...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:15:36 -0800</td>\n      <td>NaN</td>\n      <td>Pacific Time (US &amp; Canada)</td>\n      <td>really aggressive blast obnoxious entertainmen...</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>570300817074462722</td>\n      <td>0</td>\n      <td>1.0000</td>\n      <td>Can't Tell</td>\n      <td>1.0000</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>jnardino</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica and it's a really big bad thing...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:14:45 -0800</td>\n      <td>NaN</td>\n      <td>Pacific Time (US &amp; Canada)</td>\n      <td>really big bad thing</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>570300767074181121</td>\n      <td>0</td>\n      <td>1.0000</td>\n      <td>Can't Tell</td>\n      <td>0.6842</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>jnardino</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica seriously would pay $30 a fligh...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:14:33 -0800</td>\n      <td>NaN</td>\n      <td>Pacific Time (US &amp; Canada)</td>\n      <td>would pay flight seats playing really bad thin...</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>570300616901320704</td>\n      <td>1</td>\n      <td>0.6745</td>\n      <td>NaN</td>\n      <td>0.0000</td>\n      <td>Virgin America</td>\n      <td>NaN</td>\n      <td>cjmcginnis</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>@VirginAmerica yes, nearly every time I fly VX...</td>\n      <td>NaN</td>\n      <td>2015-02-24 11:13:57 -0800</td>\n      <td>San Francisco CA</td>\n      <td>Pacific Time (US &amp; Canada)</td>\n      <td>nearly every time fly vx ear worm go away</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"df['airline_sentiment'].unique()","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:19:55.618017Z","iopub.execute_input":"2022-01-06T07:19:55.618486Z","iopub.status.idle":"2022-01-06T07:19:55.625580Z","shell.execute_reply.started":"2022-01-06T07:19:55.618450Z","shell.execute_reply":"2022-01-06T07:19:55.624831Z"},"trusted":true},"execution_count":9,"outputs":[{"execution_count":9,"output_type":"execute_result","data":{"text/plain":"array([1, 0])"},"metadata":{}}]},{"cell_type":"markdown","source":"<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\"> Base SVM model with TF-IDF </h1>\n</div>\n","metadata":{}},{"cell_type":"code","source":"# Creating object of TF-IDF vectorizer\nvectorizer = TfidfVectorizer(use_idf=True, lowercase=True)\nX_tf_idf= vectorizer.fit_transform(df.cleaned_tweet)\nx_train, x_test, y_train, y_test = train_test_split(X_tf_idf, df['airline_sentiment'], random_state=42)\n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:19:55.629290Z","iopub.execute_input":"2022-01-06T07:19:55.629518Z","iopub.status.idle":"2022-01-06T07:19:55.815056Z","shell.execute_reply.started":"2022-01-06T07:19:55.629494Z","shell.execute_reply":"2022-01-06T07:19:55.814410Z"},"trusted":true},"execution_count":10,"outputs":[]},{"cell_type":"markdown","source":"","metadata":{}},{"cell_type":"markdown","source":"","metadata":{}},{"cell_type":"code","source":"\nSVM = svm.SVC( probability=True)\ns_train_accuracy, s_test_accuracy, s_train_auc, s_test_auc = check_scores(SVM,x_train, x_test, y_train, y_test)\n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:19:55.816356Z","iopub.execute_input":"2022-01-06T07:19:55.816616Z","iopub.status.idle":"2022-01-06T07:20:38.253946Z","shell.execute_reply.started":"2022-01-06T07:19:55.816584Z","shell.execute_reply":"2022-01-06T07:20:38.253164Z"},"trusted":true},"execution_count":11,"outputs":[{"name":"stdout","text":"Train confusion matrix is: \n[[6824   31]\n [ 151 1649]]\n\nTest confusion matrix is: \n[[2291   32]\n [ 296  267]]\n\n              precision    recall  f1-score   support\n\n           0       0.89      0.99      0.93      2323\n           1       0.89      0.47      0.62       563\n\n    accuracy                           0.89      2886\n   macro avg       0.89      0.73      0.78      2886\nweighted avg       0.89      0.89      0.87      2886\n\n\nTrain accuracy score:  0.9789716926632005\nTest accuracy score:  0.8863478863478863\n\nTrain ROC-AUC score:  0.9969059080962801\nTest ROC-AUC score:  0.929176839222265\n\nAre under Precision-Recall curve: 0.6194895591647333\nArea under ROC-AUC: 0.8049817790703035\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 2 Axes>","image/png":"iVBORw0KGgoAAAANSUhEUgAAAfAAAAEcCAYAAAA4KuYfAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAABWqklEQVR4nO3dd1QU19sH8C8Lq2BHiAiKWBERY4+KNUY0mESjKBp7iwqJDUuMJcYSjY2AmBgrorGg+IsdwaiANBXUKEbEACogoKDSQcp9//Bl47LLsrvM7uwuz+ecOce5U/bhyt2Hmblzrx4ABkIIIYRoFQHfARBCCCFEcZTACSGEEC1ECZwQQgjRQpTACSGEEC1ECZwQQgjRQgZ8B6AqjFHneqKb9PT0+A5B41B7J7pKVnunK3BCCCFEC1ECJ4QQQrSQzt5Cf19ltyAYY3Q7kgNUj9yQVY90i1h+1N5Vi+qRG1y0d7oCJ4QQQrQQJXBCCCFEC1ECJ4QQQrQQrwm8f//+OHPmDJKTk8EYw9SpU6s8xs7ODkFBQcjPz0dycjJWr16thkgJIdVF7Z0QbvGawOvVq4eYmBgsWLAA+fn5Ve5fv359XL58Genp6ejZsycWLFiApUuXws3NTQ3REkKqg9o7IdzitRe6v78//P39AQAHDx6scv+JEyeiTp06mDp1KgoLC/HgwQPY2NjAzc0N7u7uCn/+peR/MX6Dcn/Rvy0owB3/y0i8/bdSxxNS0/Dd3ol20NPTg1nb1qjbqKHUbT1GOKKWkRHeFhTyEB03nv4dw8l5tOo1sj59+uD69esoLPzvPy4gIAAbNmxAy5Yt8eTJE4XO98+bDPQcOVzpeHo5jcB2p8l4kfhU6XMQQqTjur0TzaBvYADTFs3RxdEBegI94P/fmLIfN1pq0tZF+gb6nJxHqxJ406ZNkZycLFaWnp4u2lZZg67snTr3mMhqxWMgFOJy1A182NisWufRBfSeMjeoHv/DdXuvahuRX1X1WMrKRIkZAB5mZeDhmwwk5WWrODLtMGHCBADV/33UqgSurMpelt9+P6La557r4oKIE39W+zzajAZ24AYN5MINGshFtSrWY32Txmhua4MBk8dBTyBAu149eIxOOxw9ehTDv1tT7fauVQk8LS0NZmbiV7vl62lpaQqfb1izNpg2bZpCx/QY4Sj2C1q3UUPYDR6A+iYm+PdWNF4+eaZwHIQQSVy3d1I9egIBTCybIS0/FyOWzkePLxxR17iRSj+zICcXKQ8fSZQbWzTFs3sPEBt2Q6WfryqvUp4D362p9nm0KoFHRERg8+bNqF27NoqKigAADg4OSElJUep5WEfjDxB19qJCx1h92FEsgTvOmyP6d1F+PtydpyHjaZLCsRBCxHHd3ol89PT0IDQ0RD0TY4z63g0tO3cCKysTJeujCTEYOOUrzj7v3l9BSImNA/Du+fiLxKdIjL6LrBcv6c5TFXhN4HXr1kXbtm0BAAKBAC1atEDnzp3x6tUrJCUlYePGjfjoo48wZMgQAO9uO6xZswYHDx7Ehg0bYG1tjeXLl2Pt2rV8/hgitevUge2Avgg5fJzvUAjROLrW3nVRl0+HYPTKJZx2JispLgbwrs9Q9ssMHF+1AS+fJeFV8nPOPqOm4jWB9+jRA0FBQaL1devWYd26dTh48CCmT58Oc3NztGnTRrQ9OzsbDg4O+PXXXxEVFYXXr19j+/btan2lpORtscztlnYdIDDQR1lJqZoiIkQ7aGN7rwlqGRnCtIUlLDvawHntimqd6+XTJMRcDUF81B0kRN9BUV7V7/sT5elBrK+g7nj/1guXnVqs+/TEnD07ZO6TmZyCnVPmIvtlhkLn1lbUOYgb8nZio7qWpKr2rsvqmRhjbZBijxDLZSY/h0lzC9w6cwH/BIfh4fVwFBcWcRyhbuOivVMCV6JBt+vdE1ad7ZDx5Bla9+iKvuOdJPa5e+kvHF5aM4Z9pC9GblACVx4l8Ko1+MAUEzatgdCwNlp27qTQseG+/8OtMxdQmJuHtPhECAQ0jUZ1cdHetaoTm6Z4HHkLjyNvAQBMrSyl7tPl0yEIO34KCdF31RgZIYS86zFu1qYVhLVrAwDa9+0Fx29nK3SOX6e5SP3+qsl/BGkaugKv5i+juXVbLDx+AAZCocS25H8eweOrGWBlZdX6DE1X069suEJX4MqjK/B3Jm1ei67Dhyp9fEpsHApz8/C/jduR9jhe6j41oR7VgW6hy6DOBm1u3RZtenRBvcaN4TBnuti2E2s24sb/zlX7MzQZNWhuUAJXXk1N4Hp6emjVvQumbNuA+iaNlT5PwG/7ELhrv1z76mI98oESuAx8NeiKfwHnZL7Cz587ozA3j9PP0STUoLlBCVx5NSmBCw1rw3H+XDT8wBRdPh2i1DkiTp7GHf/LeHb/gcKdz3SlHvlGCVwGvhp0o6Zm+O7scdQyMhSVXfM+gvPuOzn9HE1CDZoblMCVVxMSuEGtWvhq4w/oMuwThY7LTE5B3ussGNQSor6pCbZ+OQF5b7KUjkPb61FTUAKXgc8GPdRlJoa5zhKtlxQXY+uXE5DxLFnGUdqLGjQ3KIErT1cTuJ6eHj4cOhhTtm1Q6LjM5BT8PmseXqWkch6TNtajJqIELgOfDVpoWBvLz/miUdP/xnF+cO06DsxfxvlnaQJq0NygBK48XUzg9uNGw2nVUoWO+XOTOyL9zqDk7VsVRaV99aipKIHLwHeD7urogElb1omV7Z69AHERN1XyeXyiBs0NSuDK47u9c6FFJ1t86DAYNv16w7xdm6oPAPAoLBJ3/C/jcWQU3qS/UHGE72h6PWoLSuAyaEKD/tbnd7Tq1lm0nvZvAraPnaJzw6xSg+YGJXDlaUJ7rw7HeXMwZPY0ufd/kfgUHuNnoChf/UOVanI9ahNK4DJoQoNubtseC44dEBu16M9N2xF61K9a523c3ALNbW3wIvFppe9qqhM1aG5QAleeJrR3RTRuZo5mNtaY5vGzQse5O09FysM4FUUlH02qR21GCVwGTWnQ49atxEejPhet52dlY9NnY5Gfla3QeQQG+ug4sB/6jP0S7fv2BgCUFpdg3zduiIu4xWnMiqIGzQ1K4MrTlPYuj8EzJ+Ozha4KHfPDAEfkvX6jmoAUpCn1qO0ogcugKQ26vkljLL9wAoZ164rKQo/54c+N2+U63rBeXdiPG41+E8aiYZMPJLZnvXiJdZ+MqFaMjZqaoeOgfmjVrTPys7IRuGs/cl+9lvt4atDcoASuPE1p71Vp3NwCK/1PVblfwG/7kJPxCvcuX63WK1+qoAn1qAtoLHQtkJP5Cn/tOYjPF30jKrN3HoWIE3/i5dMkDJo6Ab3HjkRBdg4i/c7gwyEfw9y6DR5ej0BORgb6OI+GUf16lZ6/YlJv1a0zeo3+Aq27d0X2i5c4tmoDMpOSYVC7NmwH2KPtR91RlJeHR+E30faj7rAd2BfNbKzFztHgA1McXLic24ogpIZr1a0zvvX5vcr9fhk3Hcn/xKohIqLt6ApcDX9J6guFWHb6KExbNBeVlRQXIzMpBWatW1b7/L9Od0XLznb46MvP8UHLFmLbcl+9xsPrEej0yUAY1qtbyRnElZWW4ocBjijIzpFrf/qLnBt0Ba48TWrv0tgNHoDpnpulbstISkbAb/tw+3yAmqNSDrV3btAtdBk0rUF3/Lg/ZuzYUq1zFOXn467/X+jlVL1b5vLwXvAdYq6GyLUvNWhuUAJXnqa19/eNXrlE6pTDAHB46WrcvfSXmiOqHmrv3KAELoMmNug5e3fAundPhY/LffUaIX/4Itz3fyjIzsH2+xGcxlVWWgqBvr5EeczVYFzcsRvp8Ykyj6cGzQ1K4MrTxPYOAN9fPAlTy+ZSt/211wf+O6q+pa5pqL1zgxK4DJrYoJu2bY3FfoekJktpsjMycWWvD27876zYhAMfz5gk9kxdGUX5BYgNjcCDoFDEXg/H8vMnUKdhA6n7Lu7UR+a5qEFzgxK48jStvXce9onM4U89vpqJpJh/1BIL16i9c4MSuAya1qDLDXOdhaEuMwEAL588Q1zkLdh9PAAFOTkIOXwcDc2awLJjByTcvouwY6fwtqBA4hwCfX1svRsqVvbyyTOEn/gT0ef88fVuD1ja2gAACnPz8HfgVdy5GIj46Duw7t0TjAGJt+/ibUGh6PgPhw7G1O0/SY056cFDeH41U6xO30cNmhuUwJWnSe29TY+ucPX+rdLt25wmIzXuX5XHoSrU3rlBCVwGTWrQFbXp0RWG9evhUdgNpccsrmVkiMEzp8Cofj3EhkYiNjRC9DPXMjJE52GfoDAnF7FhkXJPF7j8/Al8YGVZ6fbsjEzUbdgQl/d44/LvB0Tl1KC5QQlceZrU3mU94lozaDhyM+V/RVMTUXvnBiVwGTSpQWsLgb4+Bk2biM8WulS57545C/Eo/AYAqkeuUAJXnia09zoNG2B9qPSe5DmZr/DjoM9U9tnqRO2dG5TAZdCEBq2tmrZtjaV/HpG5T2lxCZZ16w+A6pErlMCVx3d7/2yRKwbPmCx1m+fEWXh274FKPpcP1N65QQlcBr4btLbrOKgfZnhtrXK/guwcfNa+E7o3ba5zk7SoGyVw5fHZ3ofOnYFh33xd6faqOoFqG/re5AYlcBkogXPjy+WL0Gv0CNQyMqxy35zMV4iLuIkTazapdD5iXUUJXHl8tfdmHazhdsKn0u3ff/SxWGdRXUDfm9ygBC4DJXBu1W3UEOuuX5J7/yv7DiE2LBIJUXdUGJVuoQSuPD7au8BAH1vvhErddnLdZkSePM3ZZ2kS+t7kBiVwGSiBc0/R0eTKysrgMX4679MfagtNTOBt27bFoEGD0KRJExw5cgRPnz6FUChE06ZNkZaWhuLiYrXFIgsf7d1h7gx8KuXWuS70NJeFvje5wUV7F1S6hZAKHly7ju8/+hgBv+6Va3+BQIAvv1uk4qiIKujp6WH37t2IjY3F7t27sW7dOrRu3RoAUKtWLdy/fx/z5s3jOUp+SUvenhNm6nTyJpqFEjhRyNuCQgT+fgCLO/XBD/0/xb+3bqOugbDS/Vt374ImrazUGCHhwooVKzBjxgysXr0affr0EbsKyMvLw6lTpzB69GgeI+RX7Tp1pJY/u6+do6sR7UTTiRKl5b3Jwq4Z34Axhg4D7PH1b+5S9/vu7HH8e+s23qSm4+/Aq/gnWPpzQ6I5pk+fjgMHDmDTpk1o3LixxPb79+/j888/5yEyzdCySyeJstX9hvEQCanJ6AqccCL2egSW9xyEX6dJHwSmbc9u6DHCETN3bkW/CWOgJ6BfPU3WvHlz3Lx5s9LtBQUFqF+/vhoj0iyzd3tIlOVnZas/EFKj0bco4UxxYRESou/iaRWDVoz6fjG2/R2GeibGaoqMKCotLQ1WVpU/+ujevTuePn2qxog0RyOzJnyHQAgASuBEBXZOmSPXfmuDLmKx3yF8tsgVjZtbqDgqoohTp07BxcUFbdu2FZWV94z99NNPMWXKFJw4cULh87q4uCAhIQEFBQWIiopCv379ZO7/1Vdf4c6dO8jLy0NqaioOHz4MMzMzhT+XSxM3r5Uou3rgMA+RkJqOXiOj1yGqTVo96gkEmOG1BbYD+sp1jrw3WVg7+AuUashrSXzQpNfI6tevj6CgIFhbWyM0NBQODg64evUq6tevj549eyI6OhoDBw5EYaH8g5Q4Ozvjjz/+gKurK0JDQ+Hq6orp06fD1tYWSUlJEvvb29sjJCQES5YswenTp2FmZobffvsNr1+/xpAhQ8T2VVd7b9m5E+b9sUeifMmH9pXO1qdr6HuTG/QeuAyUwNVHVj3WN2mM4qIiuJ30gUnzZjLPc277TgQdlD0Guy7TpAQOALVr14abmxvGjh0La2trCAQCxMfHw9fXF1u3bkVRkXyz3JWLjIzEvXv3MHv2bFFZXFwc/Pz8sGLFCon9Fy9ejHnz5qFly5aismnTpsHLy0vi+bu62ru0mcZuXwzEke/WVOu82oS+N7lBCVwGSuDqI289Gtavh8lb1sOmX+9K99kwbBReP0/jMjytoWkJnEtCoRD5+fn46quv4OfnJyrfuXMn7OzsMGjQIIljevfujeDgYDg5OeH8+fMwMTHBkSNHkJWVhXHjxontq+r23tDsA3xz8HeYSHnUs+lzZ2Q8lbyDoKvoe5MbNJAL0SqFObnY67JI5jNy69491RgRqcyVK1cwePDgSrcPGjQIV65ckft8pqamMDAwQHp6ulh5eno6mjZtKvWYyMhIjB8/HkeOHMHbt2+RkZEBPT09TJ06VeZnMcakLrK2yVp+uR+JH/46KzV5f27ZDi+fPFPqvNq6KFuPtMhfj/KiBE7ULvHOPSzu1Acnftwksa12vbo8REQqGjRokMzOYk2aNMHAgQNVGkOHDh3g5eWF9evXo3v37hg2bBiaNm2K3bt3yzxOT09P6iJrW2XLNz67UCbjJmX7RqYKn1PbF2XqkRbF6lFelMAJb26cOouczFdiZSOXLuApGqKI5s2bIy8vT+79MzIyUFJSIvFHgZmZGdLSpD8y+f7773Hz5k1s27YN9+/fR2BgIFxdXTFlyhQ0aya7PwUX6ho3QpvuXSvdfmD+MpXHQIgsNBIb4VXKwziJZ+IdBvSFQKCH4qIiJET/TVOTqsmIESMwcuRI0frs2bMlensDgLGxMYYMGYIbN27Ife7i4mJER0fDwcFB7Bm4g4MDTp06JfWYOnXqoLRUfI758nWBGgYCGjRtgtRyf6/dCPI5hhIFO/ERwjVK4IRXf+09KJHAZ/26TWx94/CxyExKVmdYNZKtrS3Gjh0L4N2zuV69eqF79+5i+zDGkJeXh6CgICxapNhENe7u7jh8+DBu3ryJsLAwzJ07FxYWFvj9998BAD4+7+bVLn/Gfe7cOezduxdz585FQEAAzM3N4eHhgejoaKmvnXFt8IzJEmXfdR9If1ASjUG90BV43kCkq249Sns1pyKvSbPx5O/7Sn+GNpBVj/L8PnOptLQUkyZNwrFjxzg9r4uLC5YtWwZzc3PExMRg0aJFuH79OgDg2rVrAICPP/5YtP+3336LuXPnolWrVsjKysLVq1fx3XffISUlRey8XLf3xs3MsfLS/8TKMp4lY9NnY+U6XpfR9yY3uGjvlMDpF7HaqluPHw4djKnbf6pyv6Vd+6GspLTK/bSVJiVwbcN1e1/yvz9g3q6NWNmWkV8hPeGJ0jHqCvre5AYX7Z06sRHe3Qu8ij/kGAjDdf+vaoiGEEgkbwCUvInGoQRONMKdi4FY1m0A9sxdBPexU1EspYNQq26d4XbCh4foaqahQ4ciICAAGRkZKC4uRklJicSii7belZzu9p/gMB4iIUQ2SuBEY5QWF+NRWCRSYuOwvMcgqfs062ANowYN1BtYDTRixAhcvHgR5ubmOH78OAQCAY4dO4bjx4+jsLAQd+/exbp16/gOUyUE+voSZdLGLCCEb7wncF2YnYioxp650ns59xk7Umo54c6KFSsQHR2Nrl27Ys2ad483Dhw4gEmTJqFTp06wtLTE48ePeY6Se2atW0qUFebmIScjU/3BEFIFXhO4s7MzPD09sXHjRnTt2hXh4eHw9/eHpaWl1P3t7e1x+PBh+Pj4oGPHjvjyyy9ha2uLI0dq7gQYuuxRWCSOrVwvUf7ZQleYW7eVcgThSqdOnXD06FGUlpaK3r02MHj31mliYiJ27dqF5cuX8xmiSrT9qLtE2co+ku/CE6IJeE3gbm5uOHjwIPbt24fY2FjMnz8fqampcHFxkbp/nz59kJycDA8PDzx58gQ3btyAl5cXevXqpebIibpEnb2IV89TJcqXnKL5l1WpsLBQNFVobm4uGGNo0qSJaHtycrLYXOG64svv3fgOgRC58ZbAhUIhunfvjsDAQLHywMBA2NvbSz0mLCwM5ubm+PzzzwEAJiYmGD9+PC5evKjyeAl/Qg77Si2v05CehavKv//+C2trawBASUkJHj58CCcnJ9H2ESNG4Pnz53yFpzIVR3jLTNa9n5HoDt4SuDpnJyLaLfqcP8rKyiTK65s05iGamsHf3x/jx48X3Tb38PDAyJEjERcXh7i4OHz22WfYtWsXz1Fyq7295J28q/sP8RAJIfLRqqFU35+dqHxoxa1bt2L37t0yk7is6dkUmbqNVE7V9VjGGDweiI+9fT8mBh8Y6dbsZZry+7hhwwbs2LFD9KrY/v37UVBQgDFjxqC0tBTr16/H4cO69Rhj9m4PibJIvzPqD4QQOfGWwKs7OxEA3L9/H3l5eQgNDcWKFSskhlcsRyOxqZa66nHZmWNivYQPx9/HL+OmI/mfWJV/tjrIOzKTOpSUlODVK/GZ4o4ePYqjR4+K1vX09DTmDw5CaiLebqG/PzvR+xwcHBAeHi71GL5nJyL8kvaKzyJfb9iPG63+YGowoVCIOXPm6NRrZL1GfyFR5jVpNg+RECI/XrOeu7s7pk2bhpkzZ8LGxgYeHh4SsxOVz1AEvJudaOTIkaLJDezt7bFjxw61zU5E+PUoXPr0lU6rlkLfQKueBmksoVAIJycnLFu2DF9//TXMzc1F2wwNDbF06VI8efIEv/32m8Qf09rKxLI5nNeukCjX9clziPbj9VvvxIkTMDExwapVq0SzEw0fPhzPnj0DALRo0UJsfx8fH9SvXx/ffvsttm/fLjY7EdF9+79dii23Q6Rusx3UD/f/ClJvQDrG3NwcQUFBaNOmjehWfkFBAUaMGIHCwkIcO3YMzZs3R0REBL755hucPn2a34A5MmLpPL5DIEQpNBsZPQOvNnXXo6v3b2jTo6tE+Ypen6AoP19tcXCN79nIykda27ZtG65fv45WrVrhhx9+QE5ODkxMTBATE4Ply5dX+oiLT9Vp79Kms9WlvhVco+9NbnDR3um+I9E6v013xaxd7ujQr49YeachgxB1lsYEUJaDgwO8vb2xYsV/t5PT0tJw8uRJnDt3DqNGjdK5TmvSHr08i/mHkjfRCtTzi2ilp3/HSJR99dNqHiLRHWZmZoiMjBQrK18/ePCgziVvAHCcN0eizHuB7g0RS3QTJXCilcKPn5JavvT0UanlpGr6+vqi4VPLla9nZWXxEZLKtevdU6Is+8VLHiIhRHF0C51opbw3Wdj3zRLM+nWbWHnTNq14ikg3tG7dGj17/pfUGjZsCACwsbFBbm6uxP63bt1SW2yq0Ny2vdh6TuarSvYkRPNQJzbqjFFtfNZjxQ5IJcXF+K7bAF5iqS6+O7GVlpZKvU0ubcCW8jIDDXl9T5n2XrtOHWy8cUVsnxM/bsKNU2dVE6SOoO9NblAnNlLjrej9CTZGXql6R1Kl6dOn8x2CWn25XHK++fT4J+oPhBAlUQInWq2k6K3YuoFQyFMk2u/QoZo1ccdHoz6XKHty9x4PkRCiHOrERnRO07at+Q6BaKEXiU/5DoEQhVACJ1qt9P9ny3rf0j+P8BAJ0Sbm1m0lyv730zYpexKiuZS+hd6hQwe0bt0axsbGUh+y69pUg0S7fHf2ODaPGM93GERDNe9gLVH2763bPERCiPIUTuCtW7fGH3/8gY8++khmDzpK4ERdjq9aj/EbxAdxadLKCtZ9PkJcxE2eoiKabKjLLIkyVlbGQySEKE/hW+i7d+9Gp06dsHDhQnTr1g2tWrWSWFq3pmeQRH1unZE+fOqsX7erORKiLSoOoZr3RjcHqiG6TeEr8L59+2Ljxo3YuXOnKuIhRCnSXifTFxpAaFgbxYVFPEVFNFVDsw/E1gN+28dTJIQoT+Er8IyMDJ0dVpFor6K8fPy5yV2ifFXAnzxEoxuEQiFmzpyJP/74A4GBgejSpQuAd6OzTZ48Gc2aNeM3wGrITE4RW3+R8ISfQAipBoUT+K5duzBp0iQIBNSBnWiWW2cuSJTVa2wMY4umPESj3Ro3boxbt25hz549GDx4MAYPHgxjY2MAQHZ2NtavX49587R3Hm2Bvr7Y+sunSTxFQojyFL6FnpCQAAMDA9y7dw8+Pj5ISkpCaWmpxH4nT57kJEBC5FWUl4+H18PRob+9WPmqgD/x+EYUfp+lvQlH3TZv3gwrKyv069cPcXFxePHihWgbYwynTp2Co6Mjli/Xzpm7Kj4DL5PyHUaIplN4LHRpyboiTRgjmcZCVx9NqkejBg2wISyg0u2LO/WpdBvf+B4L/X1paWnYvXs31qxZg8aNG+Ply5cYMmQIrl27BgCYO3cuNm3aJLoq55ui7b3iGPo/DHBE3us3KotPl2hSe9dmvIyF/vHHHyt6CCFqU5CdjT83bceo7xdL3d7c1gbJ/8SqOSrt06BBAyQlVX5buXbt2kr9ke7i4oKlS5fC3NwcDx48wMKFCxEaGlrp/kKhEKtWrcLkyZNhYWGB9PR0bNu2DV5eXgp/drl6JuJ/dJSVlaEgO0fp8xHCF4VbYEhIiCriIIQzoUf90H/iOJi2aC6xbf6RvVjWtT8PUWmXx48fo0ePHti3T3rv7GHDhiEmJkahczo7O8PT0xOurq4IDQ2Fq6sr/P39YWtrW+kfC8ePH0fz5s0xe/ZsPH78GGZmZjAyMlL453nftF9+FlsXCAR0C51oJaXvcwsEAnTr1g0tW7YEADx58gTR0dFSpyMkRN02fTYWdoMHYrqn+Jd1xWefRLq9e/di27ZtCAkJQWBgIIB3t/WMjIzwww8/YOjQoZg5c6ZC53Rzc8PBgwdFfxTMnz8fn376KVxcXLBixQqJ/R0cHPDJJ5+gTZs2yMzMBAA8fVr98cpbdf2w2ucgRBMo1ZXc2dkZz549Q2RkJHx9feHr64vIyEgkJSVh7NixXMdIiFJirgbjoufvEuX0/K5qO3fuhLe3Nw4fPox///0XAODr64usrCwsW7YMv/32G3x8fOQ+n1AoRPfu3UV/DJQLDAyEvb291GO+/PJL3Lp1C25ubkhKSkJcXBw8PT1Rt25d5X8wKYJ8jnJ6PkLUReHLkREjRuDo0aOIjY3Fxo0b8fDhQwDvxkZ3cXHB0aNHUVhYiHPnznEeLCGKuvnnOQxfMFesrPsXjog6K330NvIfFxcXHDp0CM7OzmjXrh0EAgHi4+Ph6+sr87m1NKampjAwMEB6erpYeXp6OoYMGSL1mNatW6Nfv34oKiqCk5MTGjVqBC8vL1hYWMi8UJB1F7CopAQ7H94SK/Pd5A6jrTsU+GkI3WnlRnXrUeFe6Ddu3AAADBgwAEVF4iNc1apVC6GhoWCMoVevXtUKrLqoF7r6aHo9VuxxDAAbh49FZlIyD9FUTpN6oXPN3Nwcz58/x4ABA3D9+nVR+erVqzFx4kTY2NhIHBMQEID+/fujadOmyM7OBvDutnpgYCDMzMwkXm0rJ6sOa9epg003r4qVa/KbCZpI09u7tuCivSt8C93Ozg5//PGHRPIGgLdv3+Lw4cOws7NT9LSEqNWKiyfpS0iGK1euYNasWZy9JpaRkYGSkhKYmZmJlZuZmSEtLU3qMampqUhJSRElbwCiO34tWrRQKo46DeqLrb8tKFTqPIRoAoUTeH5+PkxNTSvd/sEHHyA/P79aQRHCpb0ui6SWD3VRrBNWTWJpaYndu3cjNTUVFy5cwKRJk1CvXj2lz1dcXIzo6Gg4ODiIlTs4OCA8PFzqMWFhYbCwsBB75m1t/W4aUGU7s5m1FZ9oqZaRoVLnIUQTKJzAr1y5ggULFqBfv34S2+zt7TF//nz89ddfnARHCBdiQyNxebe3RPlQl5mo07ABDxFpPmtra/Ts2ROenp7o0KEDfHx8kJ6eDj8/P4wZMwaGhoonPnd3d0ybNg0zZ86EjY0NPDw8YGFhgd9/f9fR0MfHR6xj3NGjR5GZmQlvb2/Y2trC3t4enp6eOHnyJF6+fKnUz2Vh3Uap4wjRRAo/A2/RogUiIiJgZmaG27dv49GjRwCA9u3bo1u3bkhLS4O9vT2ePXuminjlRs/A1Udb6lHas3AAWN1vGPKzsqVuUydNfgbeu3dvjBs3DmPGjIGFhQXy8vJw9uxZTJo0SaHzuLi4YNmyZTA3N0dMTAwWLVokeiZePsrb+4NFWVtbw8vLC/369cPr169x+vRpLF++HLm5uWLnlbe995swFqNX/DfIz+vUNGwYOkqhn6Gm05b2rum4aO8KJ3DgXY/S77//HsOHD4eVlRWAd7e0zp8/j82bNyMjI0PRU3KOErj6aEs9mrVphWWnJV8Zurjjd1zZK/8rUaqiyQn8fTNmzIC7uzvq1avH+5DJ5eRt7zN3bkPHQf/dPUy8cw87p8xReXy6RFvau6bjZShV4F2HlMWLF2PxYunDVRKiidLjE/EmLR2Nmop3pBo+f65GJHBNZmlpiXHjxmHcuHHo2rUrysrKcPXq1aoP1DAVR1x7+YTfO4WEVAfNCUpqlMpulxo1oGfhFZmbm2P+/PkICwtDYmIifv75ZxQUFGD+/Plo1qwZhg4dyneICqv4qKSkuJinSAipviqvwFevXg3GGH766ScwxrB69eoqT8oYw4YNGzgJkBAuMcZw9Pu1mLBpjVj5wCnjcWnnHp6i0jxBQUHo27cvBAIBoqOjsWzZMvj6+iIlJYXv0KqlmY212HpSzEOeIiGk+qp8Bl5aWioaA7m4uJimEyUStK0eaxkZYtPNa2JlKQ/j4O48laeI3tGkZ+D37t3D8ePHcfz4cSQkJKj886pL3vbuHhMpVub7w0bc/JNGjVSEtrV3TaWWZ+D6+voy1wnRNm8LCvH80WNYtG8nKkv9N57HiDTPhx/WjAk/hLVr8R0CIUqjZ+CkRgr3/VNs3dRScupRovue3L3PdwiEKE3h+9y1atVC3bp18fr1a1GZiYkJvv76azRq1Ah+fn6IioriNEhCuPa2oEBsvWWXTjxFohkSEhJQVlYGGxsblJSUICEhocqJFhhjaNu2rZoirL78EskOaxkaNh4+IYpQOIHv3bsXHTt2RI8ePQAARkZGiIyMROvW74YoXLRoET7++ONKh0ckRBNU7H1cWlLCUySaITg4GIwxlJWVia3rkjwpCbwoj4Z9JtpL4QQ+YMAAeHv/NyzlhAkT0Lp1azg6OuLu3bsICAjAqlWrMHz4cE4DJYRLCVF3xNb1NWRAEr5Mnz5d5rpu0K0/SAhR+Bm4mZkZkpKSROsjRoxAREQEAgMD8eLFC3h7e6Nr166cBkkI16RdcQsNa/MQiWaaPHmyaJRFaVq0aIHJkyerMaLqS8vPE1t/HvcvT5EQwg2FE3hOTo5oikF9fX0MHDgQly9fFm0vKChAAxoUg2g4adNImrejiS7KeXt7w97evtLtvXv3FrsTpw1KWZnYuoW19jy/J0Qahe8bRkVFYdasWbh69SpGjBiBevXq4fz586Ltbdq0qXR+X0I0Rcnbt5KF9G6rSFXv+RoZGck1JoQmof9domsUTuArV67E5cuXERUVBT09PZw8eRLR0dGi7aNGjUJYWBinQRKiCk//joFVZzvRuknzZnh27wGPEfHL0tISLVu2FK3b2Nigf//+EvsZGxtj7ty5SExMVGN01VdQKv7YJPToSZ4iIYQbCifwu3fvwsbGBvb29sjKykJISIhoW8OGDfHrr78iODiY0yAJUYdJm9ci68VLiQ5uNcX06dOxZs0aMMbAGMPKlSuxcuVKif309PRQWlqKWbNm8RCl8lLzxacgrelvHhDtp9R0otqAhlJVH22tx5+jgiCsLdlxbXGnPjxEw/9QqjY2NrC1tYWenh5OnDiBHTt2iObqfj+OvLw83LlzBy9fvlRJHMqQp36CUp8gOiNVtB4ffQe/TXNVeWy6Rlvbu6bhbTpRQnRB8j+P0Kqr5JChRg0aoCA7W8oRui02NhaxsbEA3l2Nh4SE4MmTJ/wGxaGyCu+1PwymR31Eu1XZC720tBTFxcUQCoWi9ZKSEplLMU3RR7TArhnfSC1v0clWzZFonkOHDulU8gaAZ7lZYuvFRUU8RUIIN6q8Al+3bh0YYyj5/+dF5euEaLvSkhIs7tQH2+9HiJXP/v0XXNzxO67s9eEpMvWrCdMGCyrcitTTo6kgiHbj/Rm4i4sLli5dCnNzczx48AALFy5EaGhopfsLhUKsWrUKkydPhoWFBdLT07Ft2zZ4eXmJ7UfPwNVH2+tx4fEDsOzYQaJc3c/C+XwGrq3TBpeTp34q/qF2ZqsnQg4dV2lcukjb27um0Ppn4M7OzvD09ISrqytCQ0Ph6uoKf39/2Nraio329r7jx4+jefPmmD17Nh4/fgwzMzMYGRmpOXKiSxKi70pN4OVf+OsdvsSbtHR1h6VWNXHa4JeJz/gOgZBqUfgK/LvvvsMXX3yBfv36Sd0eEhKCM2fOYPv27VWeKzIyEvfu3cPs2bNFZXFxcfDz88OKFSsk9ndwcMDJkyfRpk0bZGZmyjw3XYGrj7bXY13jRlgX4i9znyUf2qv80RHfvdC1mTz14xlzAyXvjcb2wwBH5L1+o+rQdI62t3dNwUV7V/gh0MSJExEREVHp9oiICLnGSBYKhejevTsCAwPFygMDAysdwvHLL7/ErVu34ObmhqSkJMTFxcHT0xN169ZV7Icg5D15r99Uebt8272aNbterVq1REMmlzMxMcHy5cvx888/i2Yj1CZlFa5VCnNyK9mTEO2g8C301q1b49GjR5Vuf/z4MVxcXKo8j6mpKQwMDJCeLn5rMj09HUOGDKn0s/v164eioiI4OTmhUaNG8PLygoWFBcaOHVvpZ8m6cqIOedzQhXrMKMzHoX/vVbo9OTcbzerWV2kMmlKPujhtcMXXyMq0bChYQipSOIEXFhbC3Ny80u0WFhYqGyNZIBCAMYYJEyYg+//f0/32228RGBiIJk2a4MWLF1KPo1voqqVL9WhQqxaGuszEJ7OmSGzzTXyg0o5t8t5SUwddmzZYTyB+s7GsrExj/lgiRFkK30IPDw/HjBkzpM441qhRI8yYMUOuv8ozMjJQUlICMzMzsXIzM7NKJ0NJTU1FSkqKKHkDwMOHDwG8m96QkOoqefsWFz13YVXfoVK3t/iwo5oj4oeuTRtcr7H44wCBgF4hI9pP4d/itWvXwszMDPfu3cPChQvh4OAABwcHLFq0CHfv3oWZmRnWrl1b5XmKi4sRHR0NBwcHsXIHB4dK/wAICwuDhYWF2DNva2trAMDTp08V/VEIqVRBdg7ObPGUKF9wZB8P0aifrk0b3LDJB3yHQAjnFE7g0dHR+Pzzz1FSUoLt27fD398f/v7+2LZtG4qLi/H555/j5s2bcp3L3d0d06ZNw8yZM2FjYwMPDw9YWFjg999/BwD4+PjAx+e/wTSOHj2KzMxMeHt7w9bWFvb29vD09MTJkyc1alxmohtCDtfcd4TLpw3u0qULVq5cqfXTBteqQ6+aEt2j1HvgV69eRdu2bdG9e3dRp5b4+Hjcvn1bofOcOHECJiYmWLVqFczNzRETE4Phw4fj2bN372dWvC2el5eHIUOGwMvLC7du3cLr169x+vRpLF++XJkfg5Aqnf/lV3y+SHzI1SatrPAiUbfv+OjatMGG9KYK0UG8j8SmKvQeuProcj0a1K6NzVFBYmWvnqfip2GjOf8sTXsP3MTEpNJpg6dOnYrg4GD8/fffaomlKlXVzyezpmL4grmi9X9CwrD/myVqiU3X6HJ7Vyde3gMHgMaNG2P9+vUIDQ3Fo0eP0Lt3b1H56tWrYWNjo8xpCdE4JVImvEh7nMBDJOqXmZmJc+fOiSVvAMjKysKOHTs0JnnLw9r+I7H1dh9p33vshFSk8C10KysrhIaGwsTEBPfv30ebNm1EQ5m+evUK48ePR5MmTTBv3jzOgyWEDyGHfTFg8jjRuu3AvjxGo16Ojo747LPP0LJlSwDAkydPcO7cOQQEBPAbmIJyXmaIrd88fb6SPQnRHgpfgW/ZsgV6enqwtbXF8OHDJS7vz5w5g08++YSzAAnhW0L0HYkyoWFtHiJRn9q1a+PcuXM4d+4cZs+ejU6dOqFTp06YPXs2Lly4gHPnzqFWrVoKn9fFxQUJCQkoKChAVFRUpUMyV9S3b18UFxfj/v37Cn8mAOhVGNs9K506vRLtp3ACHzJkCHbu3IknT55IHQghMTERzZs35yQ4QjTBs5h/JMpa2On2nOFr167F8OHDsW7dOpiYmMDKygpWVlZo3LgxfvzxRzg6OuLHH39U6Jzlkxdt3LgRXbt2RXh4OPz9/WFpaSnzuEaNGuHQoUO4cuWK0j9PxQuNjKRkpc9FiKZQOIEbGhri9evXlW5v1KgRysrKKt1OiLaRdrXWf9I4KXvqjvHjx+PAgQNYt24dcnJyROW5ubnYsGEDvL29MWHCBIXO6ebmhoMHD2Lfvn2IjY3F/PnzkZqaWuXQy/v374ePj4/MORgURaOwEV2gcAKPiYnBwIEDK93+5ZdfKvw6GSGa7nWq+DvPnT4ZiIFTvuIpGtUzMzNDVFRUpdujo6MlRlGURZnJi4B3t9zNzMywYcMGuT9LGomevJTAiQ5QOIF7eHjA2dkZK1asQOPGjQG8G6mpffv2OHLkCHr16gV3d3fOAyWET/oGkv09Ryydz0Mk6pGUlITBgwdXun3w4MFiQ61WRdbkRU2bNpV6jJ2dHdasWYNJkyYpdFePMSaxjHYSf+3v5MmTUvejpeqlsjqmhbt6lJfCCfzYsWNYuXIl1qxZg9jYWADApUuX8ODBA4wdOxbLli0TG7GJEF3g7jxNannf8U7qDURNDh48iDFjxmDv3r2wtbWFgYEBDAwMYGtriz179mD06NE4cOCAyj6/Vq1a8PX1xZIlS/DkyROFjtXT05NY/vzztNg+Y5ycpO5HS9VLZXVMC3f1KC+lRmLbvHkzjhw5AicnJ7Rr1w4CgQDx8fE4deqUwo2NEG2Qk5GJs1t3SFx12/Tvg7Djp3iKSnU2bdqEVq1aYcaMGZg+fbroqqD8C2b//v34+eef5T6fopMXmZubw9bWFt7e3qJZ0QQCAQQCAYqLizF8+HCxsdmrUvE7UYGLHEI0lkIJ3MjICOfPn8fhw4dx8OBBeHpKTvZAiK4KPnRMIoG379OLp2hUizGGr7/+Gp6envj8889Fwxo/ffoUFy5cQExMjELne3/yIj8/P1G5g4MDTp2S/AMoJSUFdnZ2YmWurq5wcHDAqFGjFL9QkLiqoQxOtJ9CCbygoADdu3fHsWPHVBUPIRrt1IatcFq1VLSuLzSAQF8fZaWlPEbFndq1a2PkyJFo1aoVMjIycOHCBYWutGVxd3fH4cOHcfPmTYSFhWHu3LkSkxcBwNSpU1FSUoIHDx6IHf/ixQsUFRVJlMtDD+IJXJHnjIRoKoVvoQcHB6N///7Yt69mTKtIyPvy32RJlFl1tkPibe0ZVrQy5ubmCAkJQatWrUTP4fLy8vDFF18gODi42udXdPIiLulVmP+blVECJ9pP4clMWrRogcDAQJw9exa7du2qdEAXvr0fU2WdAhijQfm5UJPqUWCgj613QsXKUh7Gwd15arXPLase5fl9rq79+/djypQp8PDwEM04uHr1aqSmpqJz584q+UyuVFU/M3/dBtsB/w2Bu++bJXgYoj2zqWmSmtTeVYmL9q7wFXhsbCwEAgHc3Nzg5uaGsrIyFBcXS3x4vXr1FD01IRqvrETyVnmzDtbQFwpRWqEdaBsHBwf4+Phg6dL/HhGkp6fj6NGjaNasGVJSUniMrnokvgQ18KKDEEUpnMCPHz+uijgI0RpR5/zR4wtHsTKbvr3wICi0kiO0Q9OmTSXm+A4NDYWenh5atGih1Qm8Yic2Rp3YiA6QO4GXd2559OgRMjMzcf78eamvfxCi6/7cuF0igc/w2orFnfrwFBE39PX1UVhYKFZWvm5oaMhHSJyp2ImNrsCJLpArgVfs3MIYQ35+PmedWwjRJoW5eSjKz0ftOnXEypvb2iD5n1ieouJG69at0bNnT9F6w4YNAQA2NjbIzc2V2P/WrVtqi6069IXiX3WUv4kukKsTmzZ2bqFObOpTE+ux73gnjF65RKws+NAxnN26Q+lz8t2JrbS0VGqH1PI/2qWVGUgZYpYPVdXP9vviE6HsmbMQj8JvqDwuXVQT27sqqK0Tmy53biFEGWHHT0kk8IFTvsIFj11a25lt+vTpfIegMnmv36CucSPRukBD/vAgpDrk+i3W6c4thCjp9oUAdPtsmFjZsjNH8fPn48C0cErdQ4cO8R2CyhTm5Ykl8EyaD5zoALkmM9Hlzi2EKOuvPQclykwtm6O9/UfqD4YopERL75IQ8j657yPpaucWQpSVnvBEannr7l0RGxqp3mAIITWO3Al87dq1WLt2rUS5l5eX2LqmdW4hRJV+GTcdi3y9xcqofw8hRB3kyrK63LmFkOpI/icWYcdPic0LbtHemseICCE1hVwJXJc7txBSXW/SXoit2/TrzVMkhJCaRK5ObISQyhnUEvIdAqmCSfNmfIdACOcogRNSTX8HXJEoq2/SmIdIuCUUCjFz5kz88ccfCAwMRJcuXQC868A6efJkNGtGSZEQPlFPM0KqSVpvdIGBvvoD4VDjxo1x9epVdOrUCenp6WjSpAmMjY0BANnZ2Vi/fj06duyI5cuX8xypcvKzsvkOgZBqoytwQjiQ++q12Pq3Prt5ioQbmzdvhpWVFfr164dOnTqJDefIGMOpU6fg6Ogo4wyaw7BeXYmywhzJV18J0TaUwAnhQL3GxmLrjZuZS00c2uKLL77Ajh07EBERIXV89MePH6NFixY8RKY4g1q1+A6BEJWgBE4IB4ryCyTKLO1seYiEGw0aNEBSUlKl22vXrq01Yz1UnAwi7/UbfgIhhGOUwAnhwI+DhkuUafOMTY8fP0aPHj0q3T5s2DDExMSoMaJqqPD/UFpSwlMghHCLEjghHHhbUCgxPWXdRg15iqb69u7di2nTpmHChAkQCN59TTDGYGRkhE2bNmHo0KH4/fffeY5SPnoC8a85aY8ECNFG2nEPjBAtIBCI9zzvPOwT3PG/zFM01bNz50507NgRhw8fRk5ODgDA19cXxsbG0NfXx6+//gofHx+eo5RPxRshlMCJrqAETghHWnfvIrbeqGkTfgLhiIuLCw4dOgRnZ2e0a9cOAoEA8fHx8PX1RWhoKN/hyU1Pr8KNRkrgREdQAieEI6HH/TBw8njRumXHDhDo66OstJTHqKonIiICERERfIdRLRX7IpRp4VzthEhDz8AJ4UhaXLxE2YdDBqk/ECJGT1DxHjo/cRDCNboCJ4QjD4Ilbys7r1uJu1KGWtV0CQkJVT4rZoyhbdu2aoqoGipcgdMzcKIrKIETwhFp7xfXrmOk/kA4EBwcLJHo9PX1YWVlhb59+yImJgZ37tzhKTrFVHwGzhjdQie6gRI4IRzyW78FY1YvEyvTEwjAtOy56/Tp0yvd9uGHHyIgIABHjhxRY0TKk3gdny7AiY6gZ+CEcOj2hQCJMku7DjxEojr37t3D7t27sXnzZr5DkUuDJh+IrWv7RDOElKMETgiHivLyJco+HPIxD5GoVnp6OmxttWOo2LJi8ZHXGluY8xQJIdyiBE6Iin08fSLfIXCqcePGmDlzJpKTk/kORS4Ve6GnxMbxFAkh3KJn4IRwLCU2Ds1srMXKTK0skfG08slBNM2VK9J7zjdq1Ag2NjaoVasWJk+erOaolFNxKNXCvDyeIiGEW3QFTgjH9rq4SZT1+EI75s4uJxAIoKenJ7YAQGJiInbu3Ak7OzscP35c4fO6uLggISEBBQUFiIqKQr9+/Srdd9SoUQgICMCLFy+QnZ2NyMhIfPHFFwp/ZsWBXFgZ9WIjuoH3K3AXFxcsXboU5ubmePDgARYuXCjXMI19+/ZFUFAQYmNj0alTJzVESoh8cjIyJcreFkhON6rJPv6Y++f2zs7O8PT0hKurK0JDQ+Hq6gp/f3/Y2tpKnbp04MCBuHr1KlatWoVXr15h4sSJ+PPPPzFo0CCFhnKVmMxEy94IIEQWxtfi7OzM3r59y2bNmsVsbGzYjh07WE5ODrO0tJR5XKNGjVh8fDy7dOkSu3//vtR93lfZeWRto0X+hepRcpnqvpFtvx8htlSnHuX5feZqMTIyYleuXGHTpk3j9LyRkZFsz549YmVxcXFs48aNcp/jxo0bbNu2bQrVT7vePcX+H+bs3cH774c2L9TeVV+P8rZ3Xm+hu7m54eDBg9i3bx9iY2Mxf/58pKamwsXFReZx+/fvh4+Pj9aP0Ux0V0F2jkRZxStBTVVQUIDu3bvDwIC7G3RCoRDdu3dHYGCgWHlgYCDs7e3lPk/9+vXx+vVrhT5b8hY6XYET3cDbN4qyDdrFxQVmZmbYsGGDqkMkRGn/hIRLlDX8wJSHSJQTHByM/v37c3Y+U1NTGBgYID09Xaw8PT0dTZs2lescrq6uaN68OQ4fPixzP8aY2HIpQPzd/KFDh0rsQ4v8i7Q6poXbepQXbwlcmQZtZ2eHNWvWYNKkSQrNKES/iPz9ItbU5f6VIInfw1O3I5WuR3WbN28eevXqhS1btqBVq1YSV7HqNnr0aGzduhUTJkzAs2fPZO5bsfPd8M+Gi22/dOmSxD60yL9Iq2NauK1HeWnHPT0AtWrVgq+vL5YsWYInT54odCz9IvL3i1iTl4ruvkpXuh7VYfLkybCysgIAxMbGwsrKCm5ubnj8+DGKioqQl5cntuTm5sp97oyMDJSUlMDMzEys3MzMDGlpaTKPdXJywuHDhzFlyhScP39e4Z/LQFhLbL20uFjhcxCiiXjrha5ogzY3N4etrS28vb3h7e0N4N2rLgKBAMXFxRg+fDguX76sltgJkcfTv2Ng1dlOrKyeiTFyMxV7hqsu3t7emDx5Mp4+fQpfX19Or/yLi4sRHR0NBwcH+Pn5icodHBxw6tSpSo8bO3YsfHx8MHXqVJn7yVJxQpm3+dr1RgAhleEtgSvaoFNSUmBnJ/5l6OrqCgcHB4waNUrhq3JCVG3/vKVYF+IvVmbZ0RYPQ8J4iki296/0ZU1moix3d3ccPnwYN2/eRFhYGObOnQsLCwv8/vvvAAAfHx8AwNSpUwEA48aNw+HDh7FkyRKEhISI/th/+/atQh3ZahmJJ/AiSuBER/D6HrgiDbqkpAQPHjwQO/7FixcoKiqSKCdEE0ibXrRD/z4am8BV7cSJEzAxMcGqVatgbm6OmJgYDB8+XPRMu0WLFmL7z507F0KhEJ6envD09BSVBwUFQZH31CtegRflS45XT4g24jWBK9qgCdF20pK6JlF1h7ldu3Zh165dUrdVTMpcDSZTq+It9IJCTs5LCN/08O6FcJ3z/hdRZZ2AGGNq6yCky6geKzd65RL0He8kWg/ctR8Bv+2Tuq+sepTn97m6SktL8fbtW7nf8GCMoV69eiqJRVGy6ueLxfMwaNoE0fp595245q0dc5lrImrv3OCivfM+lCohuiw385XYulmbVjxFIp8bN24gISGB7zA4VcvIUGydnoETXUEJnBA16jx0MOqbNEZOhcSuKXbv3o1jx47xHQanKt5CpwROdIXWvAdOiDbSr1VLomyq+0YeIqm5atepI7aubRPLEFIZSuCEqFDQwaMSZUJDQyl7ElWpeAud3gMnuoISOCEqVJCdjfioO2JlzW3b8xRNzVTxCpxuoRNdQc/ACVGxq/sPoU2PrmJldRs1RN6bLJ4ikk5fX5/vEFRC8jUySuBEN9AVOCEqlhB9V6Js3fVL6g+khhJW6IdQXFTEUySEcIsSOCEqRgOH8OuDluIDQvExuxshqkAJnBA1uOAhOfpY/4nOPERCaBASoisogROiBg+CrkuUfbl8ERp8YMpDNDVLxb4GRXk0FjrRDZTACVGDjGfJUsub29qoOZKap+IVd8nbtzxFQgi3KIETogalxcU4t81LonzE0vk8RFPDVLhjTo/Aia6gBE6ImgT5SA7q8oGVJQ+R1HSUwYluoAROiBpFnfXnO4QaR6/iJTghOoISOCFq9NfegxJlhvXqqj+QmqTCM3B6jYzoCkrghKjRq5RUibKmbVrzEEnNIfHaGCVwoiMogROiRqXFxRJlA6aM5yGSmouuwImuoAROCM86Dx3Mdwi6jR6BEx1FCZwQNXtwTXJQFz0BNUVVkbyFzk8chHCNvjUIUTO/DVslyszbteEhkpqCOrER3UQJnBA1y37xUqJs8IxJPERSU1ECJ7qBEjghGkBoZMh3CDqLJi8huooSOCE8uHvpL7F1Ye3aPEWi+yomcLqFTnQFJXBCeBAXcVNsvb19L5RRYlELqmaiKyiBE8KDYikzYqXkZfMQSQ1Ad9CJjqIETggPHkfckigLTnvKQyS6j0ZiI7qKEjghPMjJfCVR9qIwn4dIagB6Bk50FCVwQngS4XdabN1Aj5qjWlACJzqCvjEI4Un0uUti6yWsjKdIdBtNJ0p0FSVwQniS/TJToqy5bXseItFxEiOp0hU40Q2UwAnhSWZSskSZWetWPESiPi4uLkhISEBBQQGioqLQr18/mfsPGDAAUVFRKCgoQHx8PObMmVP9ICh/Ex1BCZwQDVJWpru30Z2dneHp6YmNGzeia9euCA8Ph7+/PywtLaXu37JlS1y8eBHh4eHo2rUrNm3aBC8vL4wePVqhz6WR2IiuogROCI9uXwwEALx88kznn9S6ubnh4MGD2LdvH2JjYzF//nykpqbCxcVF6v5z587F8+fPMX/+fMTGxmLfvn3w8fHBkiVLFPtg6oVOdBQlcEL49P/J5GFoBMa17ogsKROd6AKhUIju3bsjMDBQrDwwMBD29vZSj+nTp4/E/gEBAejRowcMDAzk/uy0x/FIfRwPk9pGSH0cT73Qic6QvxUQQjj37P4/6PbZMHw4ZBDMjeohIeoO3yGphKmpKQwMDJCeni5Wnp6ejiFDhkg9pmnTpvjrr78k9hcKhTA1NUVaWprU42ReYbfrjK1lExULnkiguxjcqG49UgInhEfhJ/5EgyamaGZjjdwBksOrEsVV9sybMUbPwzlA9cgNWfUob2KnBE4Ij0qLi3Hhl98AALtnL+A5GtXJyMhASUkJzMzMxMrNzMwqvZJOS0uTun9xcTEyMjJUFish2oKegRNCVK64uBjR0dFwcHAQK3dwcEB4eLjUYyIiIqTuHxUVhZKSEpXFSog2Ybq4vE/WPnzHqQsL1aPq61Ge32dNX5ydnVlRURGbOXMms7GxYR4eHiwnJ4e1aNGCAWA+Pj7Mx8dHtH/Lli1Zbm4u++WXX5iNjQ2bOXMmKyoqYqNHj1aqfrS57jRpoXpUfT0q0N75/0FUVTnUoNVX13zHoAuLridwAMzFxYUlJiaywsJCFhUVxfr37y/adu3aNXbt2jWx/QcMGMCio6NZYWEhS0hIYHPmzFG6frS97jRloXpUfT3K2971/v8fOuf9TgDUqUW1qB65Iase5fl9rsmovasP1SM3uGjv9AycEEII0UKUwAkhhBAtVCNuoROiS+j2pSRq70RX0S10QgghRMdQAieEEEK0kM7eQieEEEJ0GV2BE0IIIVqIEjghhBCihXQ6gbu4uCAhIQEFBQWIiopCv379ZO4/YMAAREVFoaCgAPHx8ZgzZ46aItVsitTjwIEDwRiTWNq3b6/GiDVL//79cebMGSQnJ4MxhqlTp1Z5jJ2dHYKCgpCfn4/k5GSsXr1aDZFqN2rv3KD2Xj3qbu+8DymnisXZ2Zm9ffuWzZo1i9nY2LAdO3awnJwcZmlpKXX/8nGXd+zYwWxsbNisWbPY27dvpY67XJMWRetx4MCBjDHGOnTowMzMzESLQCDg/Wfha3F0dGQ//fQTc3JyYnl5eWzq1Kky969fvz5LTU1lvr6+rGPHjszJyYllZ2czNzc33n8WTV2ovfNTj9TeJRc1t3f+f2BVLJGRkWzPnj1iZXFxcWzjxo1S9//5559ZXFycWNnevXtZeHg47z+LNtVjeYM2MTHhPXZNXHJycqps0HPnzmVZWVnM0NBQVLZy5UqWnJzMe/yaulB756ceqb3LXlTd3nXyFrpQKET37t0RGBgoVh4YGAh7e3upx/Tp00di/4CAAPTo0QMGBjVz2nRl6rFcVFQUnj9/jr/++guDBg1SYZS6p0+fPrh+/ToKCwtFZQEBAWjWrBlatmzJX2Aaito7N6i986M67V0nE7ipqSkMDAyQnp4uVp6eno6mTZtKPaZp06ZS9xcKhTA1NVVZrJpMmXpMTU3F3Llz4eTkhNGjR+PRo0e4cuVKlc8jyX8q+10s30bEUXvnBrV3flSnvdfMPzWJysTFxSEuLk60HhkZiZYtW2Lp0qUIDQ3lMTJCCNeovfNLJ6/AMzIyUFJSAjMzM7FyMzMzpKWlST0mLS1N6v7FxcXIyMhQWayaTJl6lObGjRto164d1+HprMp+F8u3EXHU3rlB7Z0f1WnvOpnAi4uLER0dDQcHB7FyBwcHhIeHSz0mIiJC6v5RUVEoKSlRWayaTJl6lKZLly5ITU3lOjydFRERgf79+6N27dqiMgcHB6SkpODJkyf8BaahqL1zg9o7P6rb3nnvqaeKxdnZmRUVFbGZM2cyGxsb5uHhwXJycliLFi0YAObj48N8fHxE+5e/VvLLL78wGxsbNnPmTFZUVESvlShYjwsWLGAjR45kbdu2Zba2tmzjxo2MMcZGjRrF+8/C11K3bl3WuXNn1rlzZ5aXl8dWr17NOnfuLHo1Z+PGjeyvv/4S7d+gQQOWmprKjh07xjp27MhGjRrFsrKy6DUyGQu1d37qkdq75KLm9s7/D6yqxcXFhSUmJrLCwkIWFRXF+vfvL9p27do1du3aNbH9BwwYwKKjo1lhYSFLSEhgc+bM4f1n0IRFkXpcunQpi4uLY/n5+SwzM5OFhIQwR0dH3n8GPpfyV20q8vb2ZgCYt7c3S0xMFDvGzs6OBQcHs4KCAvb8+XP2ww8/8P5zaPpC7V399UjtXXJRZ3unyUwIIYQQLaSTz8AJIYQQXUcJnBBCCNFClMAJIYQQLUQJnBBCCNFClMAJIYQQLUQJnBBCCNFClMAJ7xITE+Ht7S1at7KyAmMMU6dO5TEqQuRXU35nvb29kZiYKNe+AwcOBGMMAwcOVHFUNRcl8Bpg6tSpYIyJluLiYiQnJ8Pb2xsWFhZ8h0cIbyq2jfcXLy8vvsOrUsWY37x5g2vXrmH48OFqi+H777/HyJEj1fZ58qhYL1lZWQgKCqpWvTg6OmLNmjUcRll9NBtZDbJmzRrEx8fD0NAQvXv3xrRp09CvXz/Y2dmhqKiI7/AI4U1523jfo0ePeIpGMVeuXIG3tzf09PTQqlUruLq64ty5c/jss89w6dIlTj/r66+/hkAgft23YsUK+Pn54cyZM2LlISEhMDQ0xNu3bzmNQV7v14uVlRVcXFxw7tw5ODo6Ssx5Lo/hw4fj22+/xdq1a1UQrXIogdcgAQEBuHHjBgBg//79yMjIwPLlyzFixAicPHmS5+gI4c/7bUPbPH78GEeOHBGt//nnn7h//z4WLlzIeQJXZKIXxhivFwYV6+XUqVN4+PAhFixYoFQC10R0C70Gu379OgCgTZs2orJ27drB19cXGRkZKCgowO3bt+Hk5CRxbIMGDbBlyxbEx8ejsLAQycnJOHLkiOiWvFAoxI8//oibN2/i1atXyM/Px40bNzTuVhshsrRo0QI7d+7EP//8g7y8PLx+/Rrnzp2DnZ1dlcfWrVsXW7duRUJCAgoLC/HixQtcu3YN/fv3F9uvR48euHDhAt68eYP8/Hxcv34dgwYNUjrmmJgYvHz5Eq1atRKVDRgwAEFBQcjNzcWbN29w7tw5dOzYUeF4Kz4DZ4yhXr16mDZtmuh29bVr1wBIPgNfvHgxGGNo3bq1RMwrV64EYwxWVlYqq5fY2Fi8fPlS7PsOAL744gucPXsWSUlJKCwsxJMnT7Blyxax2cG8vb3x7bffin7m8uX9eL/66ivcvHkT+fn5ePXqFU6cOIGWLVsqHa886Aq8Biv/5Xr9+jUAwMbGBuHh4UhLS8OWLVuQm5uLUaNGwc/PD5MmTRL9NVunTh0EBwfDzs4OBw8eRFRUFExMTDB8+HC0bdsWz58/R4MGDTBnzhwcP34c3t7eMDQ0xIQJE3D69Gk4OjpyfmVASHU0bNgQJiYmYmWZmZno2bMnBgwYAD8/Pzx79gwWFhaYM2cOgoOD0bFjR5nzNe/atQvOzs749ddf8eDBAxgbG6NXr17o3Lmz6I/nAQMGICAgAHfv3sW6detQXFyMyZMnIzAwEA4ODggODlb4Z2nUqBGMjY3x77//AgAGDRqEwMBAJCYm4scff4ShoSG++eYbhIWFoWfPnnj8+LHc8VY0adIk7Nu3Dzdv3sSePXsAAOnp6VL39fX1xZYtWzBu3Dhs2rRJbNu4ceMQGRmJp0+fqqxeGjRoAGNjY4lHJdOnT0dRURF27NiBrKws9O7dG4sWLYKlpSW++uorAMDu3bthYWGBoUOHYtKkSaJjX758CQD47rvvsHHjRvj5+cHb2xvGxsb49ttvERYWhs6dO6t0jnneZ2+hRbXL1KlTGWOMDR06lJmYmLBmzZqx0aNHs/T0dFZQUMCaNWvGALCAgAAWExPDDA0NxY4PCAhgSUlJovU1a9YwxhgbM2ZMpZ8pEAhYrVq1xMqEQiG7f/8+u3z5slh5YmKiaKYeAMzKyooxxtjUqVN5rztadHspbxvS1K1bV6ItAGCtWrViBQUFbOXKlaIyab+zr169Yl5eXjI//+HDhxLtQSgUspiYGBYWFlZl/OWzXJmYmDBTU1PWo0cPdvnyZcYYY4sWLWIAWHR0NHv58iVr3Lix6Li2bduyoqIidvLkSYXilTaTVk5Ojlj7LV/KZ+UaOHCgqOz69evs7t27Yvt16NCBMcbYggULVFYv3bp1YxcvXmSMMbZ48WKxfY2MjCSO//7771lpaSlr3ry5qMzLy4sxxiT2tbS0ZG/fvmWrV68WK2/dujUrKChgP/30k8p+f+kKvAYJCAgQW09ISMCkSZOQkpICY2NjDBkyBGvXrkXdunVRt25d0X6XLl3C0KFDYW1tjbi4OIwZMwYxMTHw8/Or9LPKyspEnVeEQiHq1asHgUCAkJAQ0V+1hGiKefPm4eHDh2JlBQUFKCsrE60bGRnByMgI2dnZePToEbp37y7znFlZWejVqxcsLCzw/Plzie2dO3eGjY0NtmzZInH1f/nyZcybNw9GRkYoKCiQ+TnTpk3DtGnTROu5ublYt24dPDw80LRpU3Tr1g3bt2/Hq1evRPv8+++/OHv2LD799FMIBAKUlZVVGS8Xjh8/jp07d6J9+/aiToLjxo1DaWkpTpw4AUB19VJUVISff/4Z7u7uYvuVn0dPTw/169eHUChEaGgoBAIBunXrhuTkZJmfM3r0aAiFQvj6+orFm5WVhfv37+Pjjz+WXSnVQAm8Bin/kmrYsCGmTZuGAQMGiH5527ZtC4FAgLVr11bay7JJkyaIi4tDmzZtJHqcSjNz5kwsWrQIHTp0EOu5+v6XIiGa4NatW1I7sdWuXRvr1q3DpEmTJF65zMzMlHnOpUuXwsfHB8+ePcOdO3dw6dIlHD58GHFxcQAAa2trAMCBAwcqPYeJiUmVCeTcuXPw9PREWVkZ0tLSkJiYiMLCQgAQPaOV1qP+4cOHGDNmDExNTfHixYsq4+XCyZMn4eHhgfHjx4u+Z8aNG4fr168jNTUVAPf1IhQK0bNnT6xYsQK1a9fGu4vo/3Ts2BFbtmzBoEGDUKdOHbFtDRs2rPJnKo+3srcWKt6y5xIl8Brk/S+p06dPIzg4GMeOHUP79u1FCdbd3R0XL16UenxMTIzcnzVhwgTs27cPZ8+exebNm/HixQuUlJRg+vTpmDhxYvV/GELUwMvLCzNmzICXlxfCw8Px5s0blJWVwcPDQ+J1qor8/Pxw/fp1jBw5EkOHDsX8+fOxbNkyTJs2DceOHRMd/9133yE6OlrqOcqfscqSkpKCK1euKP7DKRgvF168eIGgoCCMGzcOa9euFV1te3h4iPZRRb1cunQJ6enp2L17N4KDg0UXIA0aNMC1a9eQl5eHlStX4t9//0VBQQGaNWsGHx+fKv+P34/3008/ldpLv6o7BdVBCbyGKisrw/LlyxEaGop58+aJ/totKSmp8ssgPj6+yl64Y8eORXx8vESv8+nTp1cvcELUaOzYsTh06BAWLVokVm5sbCxXx6T09HTs2bMHe/bsQcOGDREZGYm1a9fi2LFjoiuznJwcThKwNOWdwtq3by+xzcbGBrm5uWI/h6x4K1PxirYqx48fx759+9CpUyeMGzcOxcXFYo/jVFUv+/btg5ubGzZu3ChK4B9//DE++OADjBkzBiEhIaJ9hwwZInF8ZT9nebzPnj2TeAyjavQaWQ0WFhaG8PBwLFy4ENnZ2bh69Sq+/vprqaOzmZqaiv7t5+cHOzs7jBkzptJzl5aWAnj3XKlcq1atMGrUKA5/AkJUq7S0VOx3GADGjx+PZs2ayTxOIBCgQYMGYmVZWVlITExEo0aNAADR0dF4/Pgx3NzcUK9ePYlzvN/mlJWWlobo6GhMmTIFxsbGovLWrVtjxIgR8Pf3R1lZmVzxViYvL0/s3FU5deoU3r59i/Hjx8PZ2RlXrlwRexyhqnopKyvD9u3bYWtriy+//BKA9O8pPT09uLm5SRyfl5cHABL1cerUKZSUlOCHH36Q+rkVn+Nzia7Aa7ht27bhf//7H2bMmAEXFxeEhYXh3r172Lt3L+Lj49GkSRP06tULtra2aNeuHQBg69atcHJywrFjxzB06FBER0ejUaNGcHR0xA8//ICQkBCcPXsWTk5OOHv2LM6ePYtmzZrB1dUVjx49QteuXXn+qQmRz9mzZzFlyhRkZ2cjJiYGXbp0wbhx46p8rlm/fn2kpKTg1KlT+Pvvv5GdnY2+ffvC0dFRNEQrYwwzZ87EpUuX8M8//+DAgQNITk6GhYUFBg4cCD09PQwePLjaP8OSJUsQGBiIiIgI7N27V/QaWWFhIVauXCl3vJWJiorCkCFDsHjxYiQnJ4veH6/MmzdvEBgYiG+//RYNGjTA+vXrxbarsl4OHTqE9evX4/vvv8fp06cRFhaGjIwM+Pj4wMvLC8XFxRgzZozUPxyioqIAADt37oS/vz9KSkpw7tw5JCYmYvny5di2bRusrKxw+vRpvHnzBq1atcLIkSPh6+ur0tHbVNbFnRbNWMpflenVq5fENj09PRYXF8cSEhKYvr4+s7KyYgcOHGApKSmsqKiIJScnswsXLrDx48eLHdeoUSPm4eHBnj17xoqKilhSUhL7448/mLm5uWifJUuWsPj4eFZQUMDu37/PJk6cKHoF7f1z0WtktPC1yGobAFj9+vXZ7t27WVpaGsvNzWVBQUGsR48e7Nq1a+zatWui/Sr+zgqFQrZ582Z2+/Zt9vr1a5abm8vu37/P3NzcmL6+vthndOrUiZ04cYK9ePGCFRYWsidPnrCTJ0+yYcOGVRk/Y4zt2rWryv0GDhzIgoODWV5eHsvKymLnz59ndnZ2ou3yxivtNbK2bduyK1eusJycHMYYE9WLtNfIypeJEycyxhgrKChgDRo0kBqzqupl9erVjDHGBg8ezACwjz76iF2/fp3l5uaytLQ09uuvvzI7OzuJ7yA9PT3m7u7OUlNTWWlpKWOMMSsrK9H2ESNGsKCgIJadnc1yc3NZbGws++2335itra3Kfn/1/v8fhBBCCNEi9AycEEII0UKUwAkhhBAtRAmcEEII0UKUwAkhhBAtRAmcEEII0UKUwAkhhBAtRAmcEEII0UKUwAkhhBAtRAmcEEII0UKUwAkhhBAt9H/f7PmAfko1KAAAAABJRU5ErkJggg==\n"},"metadata":{}}]},{"cell_type":"markdown","source":"> With increase in FPR, TPR also increases.\n\n> With increase in recall, precision decreases.","metadata":{}},{"cell_type":"markdown","source":"<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\"> After optimizing the hyperparameters with TF-IDF </h1>\n</div>\n","metadata":{}},{"cell_type":"code","source":"# Tuning the hyperparameters\nparameters ={\n    \"C\":[0.1,1,10],\n    \"kernel\":['linear', 'rbf', 'sigmoid'],\n    \"gamma\":['scale', 'auto']\n}\n\n\n\nsvm_optimal = grid_search(svm.SVC(probability=True), parameters,x_train, y_train)","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:20:38.255355Z","iopub.execute_input":"2022-01-06T07:20:38.255821Z","iopub.status.idle":"2022-01-06T07:24:43.964681Z","shell.execute_reply.started":"2022-01-06T07:20:38.255786Z","shell.execute_reply":"2022-01-06T07:24:43.963913Z"},"trusted":true},"execution_count":12,"outputs":[{"name":"stdout","text":"Fitting 2 folds for each of 18 candidates, totalling 36 fits\n[CV] C=0.1, gamma=scale, kernel=linear ...............................\n","output_type":"stream"},{"name":"stderr","text":"[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n","output_type":"stream"},{"name":"stdout","text":"[CV] ................ C=0.1, gamma=scale, kernel=linear, total=   5.4s\n[CV] C=0.1, gamma=scale, kernel=linear ...............................\n","output_type":"stream"},{"name":"stderr","text":"[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    5.4s remaining:    0.0s\n","output_type":"stream"},{"name":"stdout","text":"[CV] ................ C=0.1, gamma=scale, kernel=linear, total=   5.3s\n[CV] C=0.1, gamma=scale, kernel=rbf ..................................\n[CV] ................... C=0.1, gamma=scale, kernel=rbf, total=   8.8s\n[CV] C=0.1, gamma=scale, kernel=rbf ..................................\n[CV] ................... C=0.1, gamma=scale, kernel=rbf, total=   8.4s\n[CV] C=0.1, gamma=scale, kernel=sigmoid ..............................\n[CV] ............... C=0.1, gamma=scale, kernel=sigmoid, total=   5.4s\n[CV] C=0.1, gamma=scale, kernel=sigmoid ..............................\n[CV] ............... C=0.1, gamma=scale, kernel=sigmoid, total=   5.3s\n[CV] C=0.1, gamma=auto, kernel=linear ................................\n[CV] ................. C=0.1, gamma=auto, kernel=linear, total=   5.4s\n[CV] C=0.1, gamma=auto, kernel=linear ................................\n[CV] ................. C=0.1, gamma=auto, kernel=linear, total=   5.0s\n[CV] C=0.1, gamma=auto, kernel=rbf ...................................\n[CV] .................... C=0.1, gamma=auto, kernel=rbf, total=   4.4s\n[CV] C=0.1, gamma=auto, kernel=rbf ...................................\n[CV] .................... C=0.1, gamma=auto, kernel=rbf, total=   4.2s\n[CV] C=0.1, gamma=auto, kernel=sigmoid ...............................\n[CV] ................ C=0.1, gamma=auto, kernel=sigmoid, total=   3.8s\n[CV] C=0.1, gamma=auto, kernel=sigmoid ...............................\n[CV] ................ C=0.1, gamma=auto, kernel=sigmoid, total=   3.9s\n[CV] C=1, gamma=scale, kernel=linear .................................\n[CV] .................. C=1, gamma=scale, kernel=linear, total=   5.1s\n[CV] C=1, gamma=scale, kernel=linear .................................\n[CV] .................. C=1, gamma=scale, kernel=linear, total=   5.4s\n[CV] C=1, gamma=scale, kernel=rbf ....................................\n[CV] ..................... C=1, gamma=scale, kernel=rbf, total=   9.7s\n[CV] C=1, gamma=scale, kernel=rbf ....................................\n[CV] ..................... C=1, gamma=scale, kernel=rbf, total=   9.3s\n[CV] C=1, gamma=scale, kernel=sigmoid ................................\n[CV] ................. C=1, gamma=scale, kernel=sigmoid, total=   5.0s\n[CV] C=1, gamma=scale, kernel=sigmoid ................................\n[CV] ................. C=1, gamma=scale, kernel=sigmoid, total=   4.7s\n[CV] C=1, gamma=auto, kernel=linear ..................................\n[CV] ................... C=1, gamma=auto, kernel=linear, total=   5.5s\n[CV] C=1, gamma=auto, kernel=linear ..................................\n[CV] ................... C=1, gamma=auto, kernel=linear, total=   4.9s\n[CV] C=1, gamma=auto, kernel=rbf .....................................\n[CV] ...................... C=1, gamma=auto, kernel=rbf, total=   4.6s\n[CV] C=1, gamma=auto, kernel=rbf .....................................\n[CV] ...................... C=1, gamma=auto, kernel=rbf, total=   4.4s\n[CV] C=1, gamma=auto, kernel=sigmoid .................................\n[CV] .................. C=1, gamma=auto, kernel=sigmoid, total=   3.9s\n[CV] C=1, gamma=auto, kernel=sigmoid .................................\n[CV] .................. C=1, gamma=auto, kernel=sigmoid, total=   4.1s\n[CV] C=10, gamma=scale, kernel=linear ................................\n[CV] ................. C=10, gamma=scale, kernel=linear, total=   5.4s\n[CV] C=10, gamma=scale, kernel=linear ................................\n[CV] ................. C=10, gamma=scale, kernel=linear, total=   5.7s\n[CV] C=10, gamma=scale, kernel=rbf ...................................\n[CV] .................... C=10, gamma=scale, kernel=rbf, total=  10.5s\n[CV] C=10, gamma=scale, kernel=rbf ...................................\n[CV] .................... C=10, gamma=scale, kernel=rbf, total=  10.3s\n[CV] C=10, gamma=scale, kernel=sigmoid ...............................\n[CV] ................ C=10, gamma=scale, kernel=sigmoid, total=   6.3s\n[CV] C=10, gamma=scale, kernel=sigmoid ...............................\n[CV] ................ C=10, gamma=scale, kernel=sigmoid, total=   7.0s\n[CV] C=10, gamma=auto, kernel=linear .................................\n[CV] .................. C=10, gamma=auto, kernel=linear, total=   5.3s\n[CV] C=10, gamma=auto, kernel=linear .................................\n[CV] .................. C=10, gamma=auto, kernel=linear, total=   5.5s\n[CV] C=10, gamma=auto, kernel=rbf ....................................\n[CV] ..................... C=10, gamma=auto, kernel=rbf, total=   5.5s\n[CV] C=10, gamma=auto, kernel=rbf ....................................\n[CV] ..................... C=10, gamma=auto, kernel=rbf, total=   5.7s\n[CV] C=10, gamma=auto, kernel=sigmoid ................................\n[CV] ................. C=10, gamma=auto, kernel=sigmoid, total=   4.8s\n[CV] C=10, gamma=auto, kernel=sigmoid ................................\n[CV] ................. C=10, gamma=auto, kernel=sigmoid, total=   4.7s\n","output_type":"stream"},{"name":"stderr","text":"[Parallel(n_jobs=1)]: Done  36 out of  36 | elapsed:  3.5min finished\n","output_type":"stream"},{"name":"stdout","text":"\n\nBest parameters are: \n{'C': 10, 'gamma': 'scale', 'kernel': 'rbf'}\n","output_type":"stream"}]},{"cell_type":"code","source":"so_train_accuracy, so_test_accuracy, so_train_auc, so_test_auc = check_scores(svm_optimal,x_train, x_test, y_train, y_test)","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:24:43.965825Z","iopub.execute_input":"2022-01-06T07:24:43.966078Z","iopub.status.idle":"2022-01-06T07:25:36.927423Z","shell.execute_reply.started":"2022-01-06T07:24:43.966043Z","shell.execute_reply":"2022-01-06T07:25:36.926768Z"},"trusted":true},"execution_count":13,"outputs":[{"name":"stdout","text":"Train confusion matrix is: \n[[6829   26]\n [   5 1795]]\n\nTest confusion matrix is: \n[[2272   51]\n [ 245  318]]\n\n              precision    recall  f1-score   support\n\n           0       0.90      0.98      0.94      2323\n           1       0.86      0.56      0.68       563\n\n    accuracy                           0.90      2886\n   macro avg       0.88      0.77      0.81      2886\nweighted avg       0.89      0.90      0.89      2886\n\n\nTrain accuracy score:  0.996418255343732\nTest accuracy score:  0.8974358974358975\n\nTrain ROC-AUC score:  0.9987310154793744\nTest ROC-AUC score:  0.9287425383205553\n\nAre under Precision-Recall curve: 0.6824034334763949\nArea under ROC-AUC: 0.8075544315636269\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":"> With increase in recall, preciison decreases which makes sense also.\n\n> With increase in TPR, FPR inceases.","metadata":{}},{"cell_type":"markdown","source":"<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\"> Using Multinomial Naive Bayes </h1>\n</div>\n","metadata":{}},{"cell_type":"code","source":"m_train_accuracy, m_test_accuracy, m_train_auc, m_test_auc = check_scores(MultinomialNB(),x_train, x_test, y_train, y_test)\n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:25:36.928675Z","iopub.execute_input":"2022-01-06T07:25:36.928908Z","iopub.status.idle":"2022-01-06T07:25:37.388852Z","shell.execute_reply.started":"2022-01-06T07:25:36.928876Z","shell.execute_reply":"2022-01-06T07:25:37.388174Z"},"trusted":true},"execution_count":14,"outputs":[{"name":"stdout","text":"Train confusion matrix is: \n[[6853    2]\n [1296  504]]\n\nTest confusion matrix is: \n[[2318    5]\n [ 474   89]]\n\n              precision    recall  f1-score   support\n\n           0       0.83      1.00      0.91      2323\n           1       0.95      0.16      0.27       563\n\n    accuracy                           0.83      2886\n   macro avg       0.89      0.58      0.59      2886\nweighted avg       0.85      0.83      0.78      2886\n\n\nTrain accuracy score:  0.8500288850375506\nTest accuracy score:  0.834026334026334\n\nTrain ROC-AUC score:  0.9561105438041981\nTest ROC-AUC score:  0.9013005324009118\n\nAre under Precision-Recall curve: 0.2709284627092846\nArea under ROC-AUC: 0.7374839454006489\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":"<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\"> Using Gaussian Naive Bayes\n </h1>\n</div>\n","metadata":{}},{"cell_type":"code","source":"g_train_accuracy, g_test_accuracy, g_train_auc, g_test_auc=check_scores(GaussianNB(),x_train.toarray(), x_test.toarray(), y_train, y_test)\n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:25:37.392697Z","iopub.execute_input":"2022-01-06T07:25:37.394608Z","iopub.status.idle":"2022-01-06T07:25:43.719398Z","shell.execute_reply.started":"2022-01-06T07:25:37.394570Z","shell.execute_reply":"2022-01-06T07:25:43.718722Z"},"trusted":true},"execution_count":15,"outputs":[{"name":"stdout","text":"Train confusion matrix is: \n[[5543 1312]\n [   0 1800]]\n\nTest confusion matrix is: \n[[1623  700]\n [ 181  382]]\n\n              precision    recall  f1-score   support\n\n           0       0.90      0.70      0.79      2323\n           1       0.35      0.68      0.46       563\n\n    accuracy                           0.69      2886\n   macro avg       0.63      0.69      0.63      2886\nweighted avg       0.79      0.69      0.72      2886\n\n\nTrain accuracy score:  0.8484113229347198\nTest accuracy score:  0.6947331947331947\n\nTrain ROC-AUC score:  0.9043034281546316\nTest ROC-AUC score:  0.688586755810495\n\nAre under Precision-Recall curve: 0.4644376899696049\nArea under ROC-AUC: 0.5471372315951626\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":"> It is interesting to see in Naive Bayes, we are getting linear relationship.","metadata":{}},{"cell_type":"markdown","source":"<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\"> Using AdaBoost\n </h1>\n</div>\n","metadata":{}},{"cell_type":"markdown","source":"> **base estimator here is: decision stump**","metadata":{}},{"cell_type":"code","source":"a_train_accuracy, a_test_accuracy, a_train_auc, a_test_auc=check_scores(AdaBoostClassifier(),x_train,x_test, y_train, y_test)\n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:25:43.720796Z","iopub.execute_input":"2022-01-06T07:25:43.721065Z","iopub.status.idle":"2022-01-06T07:25:47.215851Z","shell.execute_reply.started":"2022-01-06T07:25:43.721026Z","shell.execute_reply":"2022-01-06T07:25:47.215079Z"},"trusted":true},"execution_count":16,"outputs":[{"name":"stdout","text":"Train confusion matrix is: \n[[6655  200]\n [1012  788]]\n\nTest confusion matrix is: \n[[2251   72]\n [ 328  235]]\n\n              precision    recall  f1-score   support\n\n           0       0.87      0.97      0.92      2323\n           1       0.77      0.42      0.54       563\n\n    accuracy                           0.86      2886\n   macro avg       0.82      0.69      0.73      2886\nweighted avg       0.85      0.86      0.84      2886\n\n\nTrain accuracy score:  0.8599653379549393\nTest accuracy score:  0.8613998613998614\n\nTrain ROC-AUC score:  0.8689482940270687\nTest ROC-AUC score:  0.8373803091947158\n\nAre under Precision-Recall curve: 0.5402298850574713\nArea under ROC-AUC: 0.6298876361965743\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":"params = {'n_estimators': [10, 50, 100, 500],\n 'learning_rate': [0.0001, 0.001, 0.01, 0.1, 1.0],\n 'algorithm': ['SAMME', 'SAMME.R']}\n\nada_optimal_model = grid_search(AdaBoostClassifier(), params,x_train, y_train)","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:25:47.217096Z","iopub.execute_input":"2022-01-06T07:25:47.217357Z","iopub.status.idle":"2022-01-06T07:31:35.356287Z","shell.execute_reply.started":"2022-01-06T07:25:47.217328Z","shell.execute_reply":"2022-01-06T07:31:35.355511Z"},"trusted":true},"execution_count":17,"outputs":[{"name":"stdout","text":"Fitting 2 folds for each of 40 candidates, totalling 80 fits\n[CV] algorithm=SAMME, learning_rate=0.0001, n_estimators=10 ..........\n","output_type":"stream"},{"name":"stderr","text":"[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n","output_type":"stream"},{"name":"stdout","text":"[CV]  algorithm=SAMME, learning_rate=0.0001, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME, learning_rate=0.0001, n_estimators=10 ..........\n","output_type":"stream"},{"name":"stderr","text":"[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.2s remaining:    0.0s\n","output_type":"stream"},{"name":"stdout","text":"[CV]  algorithm=SAMME, learning_rate=0.0001, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME, learning_rate=0.0001, n_estimators=50 ..........\n[CV]  algorithm=SAMME, learning_rate=0.0001, n_estimators=50, total=   1.1s\n[CV] algorithm=SAMME, learning_rate=0.0001, n_estimators=50 ..........\n[CV]  algorithm=SAMME, learning_rate=0.0001, n_estimators=50, total=   1.1s\n[CV] algorithm=SAMME, learning_rate=0.0001, n_estimators=100 .........\n[CV]  algorithm=SAMME, learning_rate=0.0001, n_estimators=100, total=   2.3s\n[CV] algorithm=SAMME, learning_rate=0.0001, n_estimators=100 .........\n[CV]  algorithm=SAMME, learning_rate=0.0001, n_estimators=100, total=   2.7s\n[CV] algorithm=SAMME, learning_rate=0.0001, n_estimators=500 .........\n[CV]  algorithm=SAMME, learning_rate=0.0001, n_estimators=500, total=  11.6s\n[CV] algorithm=SAMME, learning_rate=0.0001, n_estimators=500 .........\n[CV]  algorithm=SAMME, learning_rate=0.0001, n_estimators=500, total=  12.1s\n[CV] algorithm=SAMME, learning_rate=0.001, n_estimators=10 ...........\n[CV]  algorithm=SAMME, learning_rate=0.001, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME, learning_rate=0.001, n_estimators=10 ...........\n[CV]  algorithm=SAMME, learning_rate=0.001, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME, learning_rate=0.001, n_estimators=50 ...........\n[CV]  algorithm=SAMME, learning_rate=0.001, n_estimators=50, total=   1.1s\n[CV] algorithm=SAMME, learning_rate=0.001, n_estimators=50 ...........\n[CV]  algorithm=SAMME, learning_rate=0.001, n_estimators=50, total=   1.1s\n[CV] algorithm=SAMME, learning_rate=0.001, n_estimators=100 ..........\n[CV]  algorithm=SAMME, learning_rate=0.001, n_estimators=100, total=   2.2s\n[CV] algorithm=SAMME, learning_rate=0.001, n_estimators=100 ..........\n[CV]  algorithm=SAMME, learning_rate=0.001, n_estimators=100, total=   2.2s\n[CV] algorithm=SAMME, learning_rate=0.001, n_estimators=500 ..........\n[CV]  algorithm=SAMME, learning_rate=0.001, n_estimators=500, total=  11.6s\n[CV] algorithm=SAMME, learning_rate=0.001, n_estimators=500 ..........\n[CV]  algorithm=SAMME, learning_rate=0.001, n_estimators=500, total=  12.3s\n[CV] algorithm=SAMME, learning_rate=0.01, n_estimators=10 ............\n[CV]  algorithm=SAMME, learning_rate=0.01, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME, learning_rate=0.01, n_estimators=10 ............\n[CV]  algorithm=SAMME, learning_rate=0.01, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME, learning_rate=0.01, n_estimators=50 ............\n[CV]  algorithm=SAMME, learning_rate=0.01, n_estimators=50, total=   1.1s\n[CV] algorithm=SAMME, learning_rate=0.01, n_estimators=50 ............\n[CV]  algorithm=SAMME, learning_rate=0.01, n_estimators=50, total=   1.1s\n[CV] algorithm=SAMME, learning_rate=0.01, n_estimators=100 ...........\n[CV]  algorithm=SAMME, learning_rate=0.01, n_estimators=100, total=   2.3s\n[CV] algorithm=SAMME, learning_rate=0.01, n_estimators=100 ...........\n[CV]  algorithm=SAMME, learning_rate=0.01, n_estimators=100, total=   2.2s\n[CV] algorithm=SAMME, learning_rate=0.01, n_estimators=500 ...........\n[CV]  algorithm=SAMME, learning_rate=0.01, n_estimators=500, total=  11.7s\n[CV] algorithm=SAMME, learning_rate=0.01, n_estimators=500 ...........\n[CV]  algorithm=SAMME, learning_rate=0.01, n_estimators=500, total=  12.3s\n[CV] algorithm=SAMME, learning_rate=0.1, n_estimators=10 .............\n[CV]  algorithm=SAMME, learning_rate=0.1, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME, learning_rate=0.1, n_estimators=10 .............\n[CV]  algorithm=SAMME, learning_rate=0.1, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME, learning_rate=0.1, n_estimators=50 .............\n[CV]  algorithm=SAMME, learning_rate=0.1, n_estimators=50, total=   1.5s\n[CV] algorithm=SAMME, learning_rate=0.1, n_estimators=50 .............\n[CV]  algorithm=SAMME, learning_rate=0.1, n_estimators=50, total=   1.1s\n[CV] algorithm=SAMME, learning_rate=0.1, n_estimators=100 ............\n[CV]  algorithm=SAMME, learning_rate=0.1, n_estimators=100, total=   2.3s\n[CV] algorithm=SAMME, learning_rate=0.1, n_estimators=100 ............\n[CV]  algorithm=SAMME, learning_rate=0.1, n_estimators=100, total=   2.3s\n[CV] algorithm=SAMME, learning_rate=0.1, n_estimators=500 ............\n[CV]  algorithm=SAMME, learning_rate=0.1, n_estimators=500, total=  11.5s\n[CV] algorithm=SAMME, learning_rate=0.1, n_estimators=500 ............\n[CV]  algorithm=SAMME, learning_rate=0.1, n_estimators=500, total=  12.0s\n[CV] algorithm=SAMME, learning_rate=1.0, n_estimators=10 .............\n[CV]  algorithm=SAMME, learning_rate=1.0, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME, learning_rate=1.0, n_estimators=10 .............\n[CV]  algorithm=SAMME, learning_rate=1.0, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME, learning_rate=1.0, n_estimators=50 .............\n[CV]  algorithm=SAMME, learning_rate=1.0, n_estimators=50, total=   1.2s\n[CV] algorithm=SAMME, learning_rate=1.0, n_estimators=50 .............\n[CV]  algorithm=SAMME, learning_rate=1.0, n_estimators=50, total=   1.3s\n[CV] algorithm=SAMME, learning_rate=1.0, n_estimators=100 ............\n[CV]  algorithm=SAMME, learning_rate=1.0, n_estimators=100, total=   2.6s\n[CV] algorithm=SAMME, learning_rate=1.0, n_estimators=100 ............\n[CV]  algorithm=SAMME, learning_rate=1.0, n_estimators=100, total=   2.3s\n[CV] algorithm=SAMME, learning_rate=1.0, n_estimators=500 ............\n[CV]  algorithm=SAMME, learning_rate=1.0, n_estimators=500, total=  11.6s\n[CV] algorithm=SAMME, learning_rate=1.0, n_estimators=500 ............\n[CV]  algorithm=SAMME, learning_rate=1.0, n_estimators=500, total=  12.1s\n[CV] algorithm=SAMME.R, learning_rate=0.0001, n_estimators=10 ........\n[CV]  algorithm=SAMME.R, learning_rate=0.0001, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME.R, learning_rate=0.0001, n_estimators=10 ........\n[CV]  algorithm=SAMME.R, learning_rate=0.0001, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME.R, learning_rate=0.0001, n_estimators=50 ........\n[CV]  algorithm=SAMME.R, learning_rate=0.0001, n_estimators=50, total=   1.2s\n[CV] algorithm=SAMME.R, learning_rate=0.0001, n_estimators=50 ........\n[CV]  algorithm=SAMME.R, learning_rate=0.0001, n_estimators=50, total=   1.1s\n[CV] algorithm=SAMME.R, learning_rate=0.0001, n_estimators=100 .......\n[CV]  algorithm=SAMME.R, learning_rate=0.0001, n_estimators=100, total=   2.6s\n[CV] algorithm=SAMME.R, learning_rate=0.0001, n_estimators=100 .......\n[CV]  algorithm=SAMME.R, learning_rate=0.0001, n_estimators=100, total=   2.4s\n[CV] algorithm=SAMME.R, learning_rate=0.0001, n_estimators=500 .......\n[CV]  algorithm=SAMME.R, learning_rate=0.0001, n_estimators=500, total=  12.1s\n[CV] algorithm=SAMME.R, learning_rate=0.0001, n_estimators=500 .......\n[CV]  algorithm=SAMME.R, learning_rate=0.0001, n_estimators=500, total=  12.7s\n[CV] algorithm=SAMME.R, learning_rate=0.001, n_estimators=10 .........\n[CV]  algorithm=SAMME.R, learning_rate=0.001, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME.R, learning_rate=0.001, n_estimators=10 .........\n[CV]  algorithm=SAMME.R, learning_rate=0.001, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME.R, learning_rate=0.001, n_estimators=50 .........\n[CV]  algorithm=SAMME.R, learning_rate=0.001, n_estimators=50, total=   1.2s\n[CV] algorithm=SAMME.R, learning_rate=0.001, n_estimators=50 .........\n[CV]  algorithm=SAMME.R, learning_rate=0.001, n_estimators=50, total=   1.2s\n[CV] algorithm=SAMME.R, learning_rate=0.001, n_estimators=100 ........\n[CV]  algorithm=SAMME.R, learning_rate=0.001, n_estimators=100, total=   2.7s\n[CV] algorithm=SAMME.R, learning_rate=0.001, n_estimators=100 ........\n[CV]  algorithm=SAMME.R, learning_rate=0.001, n_estimators=100, total=   2.3s\n[CV] algorithm=SAMME.R, learning_rate=0.001, n_estimators=500 ........\n[CV]  algorithm=SAMME.R, learning_rate=0.001, n_estimators=500, total=  11.9s\n[CV] algorithm=SAMME.R, learning_rate=0.001, n_estimators=500 ........\n[CV]  algorithm=SAMME.R, learning_rate=0.001, n_estimators=500, total=  12.5s\n[CV] algorithm=SAMME.R, learning_rate=0.01, n_estimators=10 ..........\n[CV]  algorithm=SAMME.R, learning_rate=0.01, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME.R, learning_rate=0.01, n_estimators=10 ..........\n[CV]  algorithm=SAMME.R, learning_rate=0.01, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME.R, learning_rate=0.01, n_estimators=50 ..........\n[CV]  algorithm=SAMME.R, learning_rate=0.01, n_estimators=50, total=   1.2s\n[CV] algorithm=SAMME.R, learning_rate=0.01, n_estimators=50 ..........\n[CV]  algorithm=SAMME.R, learning_rate=0.01, n_estimators=50, total=   1.1s\n[CV] algorithm=SAMME.R, learning_rate=0.01, n_estimators=100 .........\n[CV]  algorithm=SAMME.R, learning_rate=0.01, n_estimators=100, total=   2.4s\n[CV] algorithm=SAMME.R, learning_rate=0.01, n_estimators=100 .........\n[CV]  algorithm=SAMME.R, learning_rate=0.01, n_estimators=100, total=   2.6s\n[CV] algorithm=SAMME.R, learning_rate=0.01, n_estimators=500 .........\n[CV]  algorithm=SAMME.R, learning_rate=0.01, n_estimators=500, total=  12.0s\n[CV] algorithm=SAMME.R, learning_rate=0.01, n_estimators=500 .........\n[CV]  algorithm=SAMME.R, learning_rate=0.01, n_estimators=500, total=  12.6s\n[CV] algorithm=SAMME.R, learning_rate=0.1, n_estimators=10 ...........\n[CV]  algorithm=SAMME.R, learning_rate=0.1, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME.R, learning_rate=0.1, n_estimators=10 ...........\n[CV]  algorithm=SAMME.R, learning_rate=0.1, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME.R, learning_rate=0.1, n_estimators=50 ...........\n[CV]  algorithm=SAMME.R, learning_rate=0.1, n_estimators=50, total=   1.2s\n[CV] algorithm=SAMME.R, learning_rate=0.1, n_estimators=50 ...........\n[CV]  algorithm=SAMME.R, learning_rate=0.1, n_estimators=50, total=   1.2s\n[CV] algorithm=SAMME.R, learning_rate=0.1, n_estimators=100 ..........\n[CV]  algorithm=SAMME.R, learning_rate=0.1, n_estimators=100, total=   2.3s\n[CV] algorithm=SAMME.R, learning_rate=0.1, n_estimators=100 ..........\n[CV]  algorithm=SAMME.R, learning_rate=0.1, n_estimators=100, total=   2.7s\n[CV] algorithm=SAMME.R, learning_rate=0.1, n_estimators=500 ..........\n[CV]  algorithm=SAMME.R, learning_rate=0.1, n_estimators=500, total=  12.1s\n[CV] algorithm=SAMME.R, learning_rate=0.1, n_estimators=500 ..........\n[CV]  algorithm=SAMME.R, learning_rate=0.1, n_estimators=500, total=  12.4s\n[CV] algorithm=SAMME.R, learning_rate=1.0, n_estimators=10 ...........\n[CV]  algorithm=SAMME.R, learning_rate=1.0, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME.R, learning_rate=1.0, n_estimators=10 ...........\n[CV]  algorithm=SAMME.R, learning_rate=1.0, n_estimators=10, total=   0.2s\n[CV] algorithm=SAMME.R, learning_rate=1.0, n_estimators=50 ...........\n[CV]  algorithm=SAMME.R, learning_rate=1.0, n_estimators=50, total=   1.2s\n[CV] algorithm=SAMME.R, learning_rate=1.0, n_estimators=50 ...........\n[CV]  algorithm=SAMME.R, learning_rate=1.0, n_estimators=50, total=   1.2s\n[CV] algorithm=SAMME.R, learning_rate=1.0, n_estimators=100 ..........\n[CV]  algorithm=SAMME.R, learning_rate=1.0, n_estimators=100, total=   2.3s\n[CV] algorithm=SAMME.R, learning_rate=1.0, n_estimators=100 ..........\n[CV]  algorithm=SAMME.R, learning_rate=1.0, n_estimators=100, total=   2.6s\n[CV] algorithm=SAMME.R, learning_rate=1.0, n_estimators=500 ..........\n[CV]  algorithm=SAMME.R, learning_rate=1.0, n_estimators=500, total=  11.9s\n[CV] algorithm=SAMME.R, learning_rate=1.0, n_estimators=500 ..........\n[CV]  algorithm=SAMME.R, learning_rate=1.0, n_estimators=500, total=  12.7s\n","output_type":"stream"},{"name":"stderr","text":"[Parallel(n_jobs=1)]: Done  80 out of  80 | elapsed:  5.3min finished\n","output_type":"stream"},{"name":"stdout","text":"\n\nBest parameters are: \n{'algorithm': 'SAMME.R', 'learning_rate': 0.1, 'n_estimators': 500}\n","output_type":"stream"}]},{"cell_type":"code","source":"ao_train_accuracy, ao_test_accuracy, ao_train_auc, ao_test_auc=check_scores(ada_optimal_model,x_train,x_test, y_train, y_test)\n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:31:35.357457Z","iopub.execute_input":"2022-01-06T07:31:35.358260Z","iopub.status.idle":"2022-01-06T07:32:07.554842Z","shell.execute_reply.started":"2022-01-06T07:31:35.358218Z","shell.execute_reply":"2022-01-06T07:32:07.554186Z"},"trusted":true},"execution_count":18,"outputs":[{"name":"stdout","text":"Train confusion matrix is: \n[[6761   94]\n [1043  757]]\n\nTest confusion matrix is: \n[[2278   45]\n [ 340  223]]\n\n              precision    recall  f1-score   support\n\n           0       0.87      0.98      0.92      2323\n           1       0.83      0.40      0.54       563\n\n    accuracy                           0.87      2886\n   macro avg       0.85      0.69      0.73      2886\nweighted avg       0.86      0.87      0.85      2886\n\n\nTrain accuracy score:  0.868630849220104\nTest accuracy score:  0.8665973665973666\n\nTrain ROC-AUC score:  0.9218550936056407\nTest ROC-AUC score:  0.8779339205061136\n\nAre under Precision-Recall curve: 0.5367027677496993\nArea under ROC-AUC: 0.7021179688268547\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":"<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\"> Bagging with MultinomialNB\n </h1>\n</div>\n","metadata":{}},{"cell_type":"code","source":"\n\nkfold = model_selection.KFold(n_splits = 3)\n  \n# bagging classifier\nmodel = BaggingClassifier(base_estimator = MultinomialNB(),\n                          n_estimators = 100)\n\nb_train_accuracy, b_test_accuracy, b_train_auc, b_test_auc= check_scores(model,x_train,x_test, y_train, y_test)\n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:32:07.556017Z","iopub.execute_input":"2022-01-06T07:32:07.556354Z","iopub.status.idle":"2022-01-06T07:32:09.342833Z","shell.execute_reply.started":"2022-01-06T07:32:07.556312Z","shell.execute_reply":"2022-01-06T07:32:09.342190Z"},"trusted":true},"execution_count":19,"outputs":[{"name":"stdout","text":"Train confusion matrix is: \n[[6853    2]\n [1318  482]]\n\nTest confusion matrix is: \n[[2319    4]\n [ 475   88]]\n\n              precision    recall  f1-score   support\n\n           0       0.83      1.00      0.91      2323\n           1       0.96      0.16      0.27       563\n\n    accuracy                           0.83      2886\n   macro avg       0.89      0.58      0.59      2886\nweighted avg       0.85      0.83      0.78      2886\n\n\nTrain accuracy score:  0.8474870017331022\nTest accuracy score:  0.834026334026334\n\nTrain ROC-AUC score:  0.9525061188102764\nTest ROC-AUC score:  0.8998424130002777\n\nAre under Precision-Recall curve: 0.2687022900763359\nArea under ROC-AUC: 0.7345749814146296\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":"<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\">Using Random Forest\n </h1>\n</div>\n\n","metadata":{}},{"cell_type":"code","source":"r_train_accuracy, r_test_accuracy, r_train_auc, r_test_auc= check_scores(RandomForestClassifier(random_state=0).fit(x_train, y_train), x_train,x_test,y_train,y_test)\n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:32:09.344047Z","iopub.execute_input":"2022-01-06T07:32:09.344510Z","iopub.status.idle":"2022-01-06T07:32:36.268949Z","shell.execute_reply.started":"2022-01-06T07:32:09.344472Z","shell.execute_reply":"2022-01-06T07:32:36.268210Z"},"trusted":true},"execution_count":20,"outputs":[{"name":"stdout","text":"Train confusion matrix is: \n[[6829   26]\n [   5 1795]]\n\nTest confusion matrix is: \n[[2215  108]\n [ 238  325]]\n\n              precision    recall  f1-score   support\n\n           0       0.90      0.95      0.93      2323\n           1       0.75      0.58      0.65       563\n\n    accuracy                           0.88      2886\n   macro avg       0.83      0.77      0.79      2886\nweighted avg       0.87      0.88      0.87      2886\n\n\nTrain accuracy score:  0.996418255343732\nTest accuracy score:  0.8801108801108801\n\nTrain ROC-AUC score:  0.9982442661479861\nTest ROC-AUC score:  0.8956867344777572\n\nAre under Precision-Recall curve: 0.6526104417670683\nArea under ROC-AUC: 0.7441899264879837\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":"<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\">Using LSTM\n </h1>\n</div>\n\n","metadata":{}},{"cell_type":"code","source":"corpus = [df['cleaned_tweet'][i] for i in range( len(df))]\n\nvoc_size=5000\n\nonehot_=[one_hot(words,voc_size)for words in corpus] \n\nmax_sent_length=max([len(i) for i in corpus])\n\nembedded_docs=pad_sequences(onehot_,padding='pre',maxlen=max_sent_length)\n    \nembedding_vector_features=40\nmodel=Sequential()\nmodel.add(Embedding(voc_size,embedding_vector_features,input_length=max_sent_length))\nmodel.add(Dropout(0.3))\nmodel.add(LSTM(100))\nmodel.add(Dropout(0.3))\nmodel.add(Dense(1,activation='sigmoid'))\nmodel.compile(loss='binary_crossentropy',optimizer='adam',metrics=['accuracy'])\n\nX_final=np.array(embedded_docs)\ny_final=np.array(df['airline_sentiment'])\nX_final.shape,y_final.shape\n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:32:36.271159Z","iopub.execute_input":"2022-01-06T07:32:36.271688Z","iopub.status.idle":"2022-01-06T07:32:38.849315Z","shell.execute_reply.started":"2022-01-06T07:32:36.271652Z","shell.execute_reply":"2022-01-06T07:32:38.848620Z"},"trusted":true},"execution_count":21,"outputs":[{"name":"stderr","text":"2022-01-06 07:32:36.602285: I tensorflow/compiler/jit/xla_cpu_device.cc:41] Not creating XLA devices, tf_xla_enable_xla_devices not set\n2022-01-06 07:32:36.606309: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcuda.so.1\n2022-01-06 07:32:36.657305: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:941] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-01-06 07:32:36.658367: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1720] Found device 0 with properties: \npciBusID: 0000:00:04.0 name: Tesla P100-PCIE-16GB computeCapability: 6.0\ncoreClock: 1.3285GHz coreCount: 56 deviceMemorySize: 15.90GiB deviceMemoryBandwidth: 681.88GiB/s\n2022-01-06 07:32:36.658482: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0\n2022-01-06 07:32:36.685884: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublas.so.11\n2022-01-06 07:32:36.686073: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublasLt.so.11\n2022-01-06 07:32:36.708032: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcufft.so.10\n2022-01-06 07:32:36.716806: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcurand.so.10\n2022-01-06 07:32:36.741650: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusolver.so.10\n2022-01-06 07:32:36.748864: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusparse.so.11\n2022-01-06 07:32:36.752676: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudnn.so.8\n2022-01-06 07:32:36.752918: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:941] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-01-06 07:32:36.753983: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:941] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-01-06 07:32:36.755817: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1862] Adding visible gpu devices: 0\n2022-01-06 07:32:36.757898: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 AVX512F FMA\nTo enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n2022-01-06 07:32:36.758164: I tensorflow/compiler/jit/xla_gpu_device.cc:99] Not creating XLA devices, tf_xla_enable_xla_devices not set\n2022-01-06 07:32:36.758412: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:941] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-01-06 07:32:36.759322: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1720] Found device 0 with properties: \npciBusID: 0000:00:04.0 name: Tesla P100-PCIE-16GB computeCapability: 6.0\ncoreClock: 1.3285GHz coreCount: 56 deviceMemorySize: 15.90GiB deviceMemoryBandwidth: 681.88GiB/s\n2022-01-06 07:32:36.759425: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0\n2022-01-06 07:32:36.759471: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublas.so.11\n2022-01-06 07:32:36.759506: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublasLt.so.11\n2022-01-06 07:32:36.759540: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcufft.so.10\n2022-01-06 07:32:36.759576: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcurand.so.10\n2022-01-06 07:32:36.759612: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusolver.so.10\n2022-01-06 07:32:36.759648: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusparse.so.11\n2022-01-06 07:32:36.759685: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudnn.so.8\n2022-01-06 07:32:36.759806: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:941] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-01-06 07:32:36.760845: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:941] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-01-06 07:32:36.761750: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1862] Adding visible gpu devices: 0\n2022-01-06 07:32:36.763126: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0\n2022-01-06 07:32:38.247897: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1261] Device interconnect StreamExecutor with strength 1 edge matrix:\n2022-01-06 07:32:38.247944: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1267]      0 \n2022-01-06 07:32:38.247954: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1280] 0:   N \n2022-01-06 07:32:38.250100: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:941] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-01-06 07:32:38.250843: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:941] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-01-06 07:32:38.251488: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:941] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-01-06 07:32:38.252040: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1406] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 14957 MB memory) -> physical GPU (device: 0, name: Tesla P100-PCIE-16GB, pci bus id: 0000:00:04.0, compute capability: 6.0)\n","output_type":"stream"},{"execution_count":21,"output_type":"execute_result","data":{"text/plain":"((11541, 124), (11541,))"},"metadata":{}}]},{"cell_type":"code","source":"X_train, X_test, y_train, y_test = train_test_split(X_final, y_final, test_size=0.33, random_state=42)\nmodel.fit(X_train,y_train,validation_data=(X_test,y_test),epochs=10,batch_size=64)\n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:32:38.850553Z","iopub.execute_input":"2022-01-06T07:32:38.850812Z","iopub.status.idle":"2022-01-06T07:33:03.491439Z","shell.execute_reply.started":"2022-01-06T07:32:38.850776Z","shell.execute_reply":"2022-01-06T07:33:03.490756Z"},"trusted":true},"execution_count":22,"outputs":[{"name":"stderr","text":"2022-01-06 07:32:38.903020: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:116] None of the MLIR optimization passes are enabled (registered 2)\n2022-01-06 07:32:38.913133: I tensorflow/core/platform/profile_utils/cpu_utils.cc:112] CPU Frequency: 2000170000 Hz\n","output_type":"stream"},{"name":"stdout","text":"Epoch 1/10\n","output_type":"stream"},{"name":"stderr","text":"2022-01-06 07:32:40.600036: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublas.so.11\n2022-01-06 07:32:41.447746: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublasLt.so.11\n2022-01-06 07:32:41.501725: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudnn.so.8\n","output_type":"stream"},{"name":"stdout","text":"121/121 [==============================] - 10s 19ms/step - loss: 0.5460 - accuracy: 0.7747 - val_loss: 0.3390 - val_accuracy: 0.8519\nEpoch 2/10\n121/121 [==============================] - 2s 13ms/step - loss: 0.2935 - accuracy: 0.8775 - val_loss: 0.2888 - val_accuracy: 0.8892\nEpoch 3/10\n121/121 [==============================] - 2s 14ms/step - loss: 0.1997 - accuracy: 0.9193 - val_loss: 0.2902 - val_accuracy: 0.8887\nEpoch 4/10\n121/121 [==============================] - 2s 14ms/step - loss: 0.1589 - accuracy: 0.9391 - val_loss: 0.3124 - val_accuracy: 0.8792\nEpoch 5/10\n121/121 [==============================] - 2s 13ms/step - loss: 0.1253 - accuracy: 0.9549 - val_loss: 0.3389 - val_accuracy: 0.8779\nEpoch 6/10\n121/121 [==============================] - 2s 13ms/step - loss: 0.1113 - accuracy: 0.9613 - val_loss: 0.3745 - val_accuracy: 0.8769\nEpoch 7/10\n121/121 [==============================] - 2s 13ms/step - loss: 0.0898 - accuracy: 0.9671 - val_loss: 0.3949 - val_accuracy: 0.8711\nEpoch 8/10\n121/121 [==============================] - 2s 13ms/step - loss: 0.0858 - accuracy: 0.9694 - val_loss: 0.4239 - val_accuracy: 0.8703\nEpoch 9/10\n121/121 [==============================] - 2s 13ms/step - loss: 0.0758 - accuracy: 0.9712 - val_loss: 0.4847 - val_accuracy: 0.8658\nEpoch 10/10\n121/121 [==============================] - 2s 15ms/step - loss: 0.0651 - accuracy: 0.9763 - val_loss: 0.5238 - val_accuracy: 0.8635\n","output_type":"stream"},{"execution_count":22,"output_type":"execute_result","data":{"text/plain":"<tensorflow.python.keras.callbacks.History at 0x7f76685ec1d0>"},"metadata":{}}]},{"cell_type":"code","source":"y_test_pred=model.predict_classes(X_test)\ny_train_pred=model.predict_classes(X_train)\n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:33:03.492662Z","iopub.execute_input":"2022-01-06T07:33:03.492996Z","iopub.status.idle":"2022-01-06T07:33:04.881827Z","shell.execute_reply.started":"2022-01-06T07:33:03.492960Z","shell.execute_reply":"2022-01-06T07:33:04.881099Z"},"trusted":true},"execution_count":23,"outputs":[{"name":"stderr","text":"/opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/sequential.py:450: UserWarning: `model.predict_classes()` is deprecated and will be removed after 2021-01-01. Please use instead:* `np.argmax(model.predict(x), axis=-1)`,   if your model does multi-class classification   (e.g. if it uses a `softmax` last-layer activation).* `(model.predict(x) > 0.5).astype(\"int32\")`,   if your model does binary classification   (e.g. if it uses a `sigmoid` last-layer activation).\n  warnings.warn('`model.predict_classes()` is deprecated and '\n","output_type":"stream"}]},{"cell_type":"code","source":"test_acc_lstm = accuracy_score(y_test,y_test_pred)\ntrain_acc_lstm = accuracy_score(y_train,y_train_pred)\ntest_roc_lstm = roc_auc_score(y_test,y_test_pred)\ntrain_roc_lstm = roc_auc_score(y_train,y_train_pred)\n","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:33:04.885021Z","iopub.execute_input":"2022-01-06T07:33:04.885274Z","iopub.status.idle":"2022-01-06T07:33:04.896350Z","shell.execute_reply.started":"2022-01-06T07:33:04.885246Z","shell.execute_reply":"2022-01-06T07:33:04.895690Z"},"trusted":true},"execution_count":24,"outputs":[]},{"cell_type":"markdown","source":"<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\">Final Results\n </h1>\n</div>\n\n","metadata":{}},{"cell_type":"code","source":"data = [('Random Forest', r_train_accuracy, r_test_accuracy, r_train_auc, r_test_auc),\n ('MultinomialNB',m_train_accuracy, m_test_accuracy, m_train_auc, m_test_auc  ),\n('Bagged MultinomialNB',b_train_accuracy, b_test_accuracy, b_train_auc, b_test_auc ),\n ('AdaBoost',a_train_accuracy, a_test_accuracy, a_train_auc, a_test_auc ),\n('AdaBoost Optimized',ao_train_accuracy, ao_test_accuracy, ao_train_auc, ao_test_auc),\n('Gaussian Naive Bayes',g_train_accuracy, g_test_accuracy, g_train_auc, g_test_auc),\n('SVM', s_train_accuracy, s_test_accuracy, s_train_auc, s_test_auc),\n('SVM Optimized', so_train_accuracy, so_test_accuracy, so_train_auc, so_test_auc),\n('LSTM',train_acc_lstm, test_acc_lstm, train_roc_lstm, test_roc_lstm )]\n\n\nScores_ =pd.DataFrame(data = data, columns=['Model Name','Train Accuracy', 'Test Accuracy', 'Train ROC', 'Test ROC'])\nScores_.set_index('Model Name', inplace = True)\n\nScores_","metadata":{"execution":{"iopub.status.busy":"2022-01-06T07:33:04.897499Z","iopub.execute_input":"2022-01-06T07:33:04.897820Z","iopub.status.idle":"2022-01-06T07:33:04.914992Z","shell.execute_reply.started":"2022-01-06T07:33:04.897783Z","shell.execute_reply":"2022-01-06T07:33:04.914291Z"},"trusted":true},"execution_count":25,"outputs":[{"execution_count":25,"output_type":"execute_result","data":{"text/plain":"                      Train Accuracy  Test Accuracy  Train ROC  Test ROC\nModel Name                                                              \nRandom Forest               0.996418       0.880111   0.998244  0.895687\nMultinomialNB               0.850029       0.834026   0.956111  0.901301\nBagged MultinomialNB        0.847487       0.834026   0.952506  0.899842\nAdaBoost                    0.859965       0.861400   0.868948  0.837380\nAdaBoost Optimized          0.868631       0.866597   0.921855  0.877934\nGaussian Naive Bayes        0.848411       0.694733   0.904303  0.688587\nSVM                         0.978972       0.886348   0.996906  0.929177\nSVM Optimized               0.996418       0.897436   0.998731  0.928743\nLSTM                        0.984868       0.863481   0.980694  0.787352","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>Train Accuracy</th>\n      <th>Test Accuracy</th>\n      <th>Train ROC</th>\n      <th>Test ROC</th>\n    </tr>\n    <tr>\n      <th>Model Name</th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>Random Forest</th>\n      <td>0.996418</td>\n      <td>0.880111</td>\n      <td>0.998244</td>\n      <td>0.895687</td>\n    </tr>\n    <tr>\n      <th>MultinomialNB</th>\n      <td>0.850029</td>\n      <td>0.834026</td>\n      <td>0.956111</td>\n      <td>0.901301</td>\n    </tr>\n    <tr>\n      <th>Bagged MultinomialNB</th>\n      <td>0.847487</td>\n      <td>0.834026</td>\n      <td>0.952506</td>\n      <td>0.899842</td>\n    </tr>\n    <tr>\n      <th>AdaBoost</th>\n      <td>0.859965</td>\n      <td>0.861400</td>\n      <td>0.868948</td>\n      <td>0.837380</td>\n    </tr>\n    <tr>\n      <th>AdaBoost Optimized</th>\n      <td>0.868631</td>\n      <td>0.866597</td>\n      <td>0.921855</td>\n      <td>0.877934</td>\n    </tr>\n    <tr>\n      <th>Gaussian Naive Bayes</th>\n      <td>0.848411</td>\n      <td>0.694733</td>\n      <td>0.904303</td>\n      <td>0.688587</td>\n    </tr>\n    <tr>\n      <th>SVM</th>\n      <td>0.978972</td>\n      <td>0.886348</td>\n      <td>0.996906</td>\n      <td>0.929177</td>\n    </tr>\n    <tr>\n      <th>SVM Optimized</th>\n      <td>0.996418</td>\n      <td>0.897436</td>\n      <td>0.998731</td>\n      <td>0.928743</td>\n    </tr>\n    <tr>\n      <th>LSTM</th>\n      <td>0.984868</td>\n      <td>0.863481</td>\n      <td>0.980694</td>\n      <td>0.787352</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"markdown","source":"\n<div style=\"color:white;\n           display:fill;\n           border-radius:5px;\n           font-size:110%;\n           font-family:cursive;\n           letter-spacing:0.5px;\n           background-color:powderblue;\n           color:Black;\n           font-family:cursive\n           \">\n<h1 style=\"text-align:center;font-weight: bold\">Conclusion\n </h1>\n</div>\n\n**Most of the models are doing pretty well here.**","metadata":{}}]}