notebook/Credit_Card_Fraud_Detection.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"Credit_Card_Fraud_Detection.ipynb","provenance":[],"collapsed_sections":[],"machine_shape":"hm"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"},"accelerator":"GPU","gpuClass":"standard"},"cells":[{"cell_type":"markdown","source":["# Credit Card Fraud Detection "],"metadata":{"id":"QaXKOjAiDTtM"}},{"cell_type":"markdown","source":["## Introduction:\n","* With the provided information, build a model to predict whether this customer will commit fraud when using a credit card or not?\n","* It is important for credit card issuers to detect customers who commit credit card fraud so that when they buy products and services, they will not be charged for those products and services.\n","* This dataset contains transaction information with 492 TH frauds out of a total of 284,807 transactions. This is a severely imbalanced dataset with the positive class (fraud) accounting for 0.172% of all transactions. Therefore, we should not use metrics `Accuracy` to evaluate the model because the number of zero labels is too much, instead, we should focus on F1 Score, Precision/Recall, ROC_AUC or Confusion matrix to have more accurate assessment of the performance of the models"],"metadata":{"id":"6-i7_b6tD7vz"}},{"cell_type":"code","source":["import pandas as pd\n","import numpy as np\n","import matplotlib.pyplot as plt\n","import seaborn as sns\n","from sklearn import preprocessing \n","from sklearn.preprocessing import LabelEncoder\n","from imblearn.over_sampling import SMOTE\n","%matplotlib inline\n","from sklearn.ensemble import RandomForestClassifier\n","from sklearn.linear_model import LogisticRegression\n","from sklearn import datasets, linear_model, metrics\n","from sklearn.ensemble import VotingClassifier\n","from sklearn.model_selection import StratifiedKFold\n","from sklearn.model_selection import cross_val_score\n","from sklearn.metrics import accuracy_score,precision_score, recall_score, f1_score, confusion_matrix, classification_report, precision_recall_curve\n","from sklearn.preprocessing import StandardScaler\n","from sklearn.model_selection import train_test_split\n","import warnings\n","warnings.filterwarnings('ignore')"],"metadata":{"id":"QnKDx3EX8oHk","executionInfo":{"status":"ok","timestamp":1659231665788,"user_tz":-420,"elapsed":315,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":62,"outputs":[]},{"cell_type":"markdown","source":["##Load Data"],"metadata":{"id":"w4NWaOn2s_A0"}},{"cell_type":"code","source":["df=pd.read_csv('dataset/creditcard.csv')\n","# Converting all column names to lower case\n","df.columns = df.columns.str.lower()\n","df.head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":299},"id":"KUh5wUQCrahJ","outputId":"a30b8f73-7a9b-4d73-a5eb-5e14e731a10b","executionInfo":{"status":"ok","timestamp":1659223922683,"user_tz":-420,"elapsed":2816,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":4,"outputs":[{"output_type":"execute_result","data":{"text/plain":["   time        v1        v2        v3        v4        v5        v6        v7  \\\n","0   0.0 -1.359807 -0.072781  2.536347  1.378155 -0.338321  0.462388  0.239599   \n","1   0.0  1.191857  0.266151  0.166480  0.448154  0.060018 -0.082361 -0.078803   \n","2   1.0 -1.358354 -1.340163  1.773209  0.379780 -0.503198  1.800499  0.791461   \n","3   1.0 -0.966272 -0.185226  1.792993 -0.863291 -0.010309  1.247203  0.237609   \n","4   2.0 -1.158233  0.877737  1.548718  0.403034 -0.407193  0.095921  0.592941   \n","\n","         v8        v9  ...       v21       v22       v23       v24       v25  \\\n","0  0.098698  0.363787  ... -0.018307  0.277838 -0.110474  0.066928  0.128539   \n","1  0.085102 -0.255425  ... -0.225775 -0.638672  0.101288 -0.339846  0.167170   \n","2  0.247676 -1.514654  ...  0.247998  0.771679  0.909412 -0.689281 -0.327642   \n","3  0.377436 -1.387024  ... -0.108300  0.005274 -0.190321 -1.175575  0.647376   \n","4 -0.270533  0.817739  ... -0.009431  0.798278 -0.137458  0.141267 -0.206010   \n","\n","        v26       v27       v28  amount  class  \n","0 -0.189115  0.133558 -0.021053  149.62      0  \n","1  0.125895 -0.008983  0.014724    2.69      0  \n","2 -0.139097 -0.055353 -0.059752  378.66      0  \n","3 -0.221929  0.062723  0.061458  123.50      0  \n","4  0.502292  0.219422  0.215153   69.99      0  \n","\n","[5 rows x 31 columns]"],"text/html":["\n","  <div id=\"df-64766cb3-5f58-4ab4-ae53-96500bb40306\">\n","    <div class=\"colab-df-container\">\n","      <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>time</th>\n","      <th>v1</th>\n","      <th>v2</th>\n","      <th>v3</th>\n","      <th>v4</th>\n","      <th>v5</th>\n","      <th>v6</th>\n","      <th>v7</th>\n","      <th>v8</th>\n","      <th>v9</th>\n","      <th>...</th>\n","      <th>v21</th>\n","      <th>v22</th>\n","      <th>v23</th>\n","      <th>v24</th>\n","      <th>v25</th>\n","      <th>v26</th>\n","      <th>v27</th>\n","      <th>v28</th>\n","      <th>amount</th>\n","      <th>class</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>0.0</td>\n","      <td>-1.359807</td>\n","      <td>-0.072781</td>\n","      <td>2.536347</td>\n","      <td>1.378155</td>\n","      <td>-0.338321</td>\n","      <td>0.462388</td>\n","      <td>0.239599</td>\n","      <td>0.098698</td>\n","      <td>0.363787</td>\n","      <td>...</td>\n","      <td>-0.018307</td>\n","      <td>0.277838</td>\n","      <td>-0.110474</td>\n","      <td>0.066928</td>\n","      <td>0.128539</td>\n","      <td>-0.189115</td>\n","      <td>0.133558</td>\n","      <td>-0.021053</td>\n","      <td>149.62</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>0.0</td>\n","      <td>1.191857</td>\n","      <td>0.266151</td>\n","      <td>0.166480</td>\n","      <td>0.448154</td>\n","      <td>0.060018</td>\n","      <td>-0.082361</td>\n","      <td>-0.078803</td>\n","      <td>0.085102</td>\n","      <td>-0.255425</td>\n","      <td>...</td>\n","      <td>-0.225775</td>\n","      <td>-0.638672</td>\n","      <td>0.101288</td>\n","      <td>-0.339846</td>\n","      <td>0.167170</td>\n","      <td>0.125895</td>\n","      <td>-0.008983</td>\n","      <td>0.014724</td>\n","      <td>2.69</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>1.0</td>\n","      <td>-1.358354</td>\n","      <td>-1.340163</td>\n","      <td>1.773209</td>\n","      <td>0.379780</td>\n","      <td>-0.503198</td>\n","      <td>1.800499</td>\n","      <td>0.791461</td>\n","      <td>0.247676</td>\n","      <td>-1.514654</td>\n","      <td>...</td>\n","      <td>0.247998</td>\n","      <td>0.771679</td>\n","      <td>0.909412</td>\n","      <td>-0.689281</td>\n","      <td>-0.327642</td>\n","      <td>-0.139097</td>\n","      <td>-0.055353</td>\n","      <td>-0.059752</td>\n","      <td>378.66</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>1.0</td>\n","      <td>-0.966272</td>\n","      <td>-0.185226</td>\n","      <td>1.792993</td>\n","      <td>-0.863291</td>\n","      <td>-0.010309</td>\n","      <td>1.247203</td>\n","      <td>0.237609</td>\n","      <td>0.377436</td>\n","      <td>-1.387024</td>\n","      <td>...</td>\n","      <td>-0.108300</td>\n","      <td>0.005274</td>\n","      <td>-0.190321</td>\n","      <td>-1.175575</td>\n","      <td>0.647376</td>\n","      <td>-0.221929</td>\n","      <td>0.062723</td>\n","      <td>0.061458</td>\n","      <td>123.50</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>2.0</td>\n","      <td>-1.158233</td>\n","      <td>0.877737</td>\n","      <td>1.548718</td>\n","      <td>0.403034</td>\n","      <td>-0.407193</td>\n","      <td>0.095921</td>\n","      <td>0.592941</td>\n","      <td>-0.270533</td>\n","      <td>0.817739</td>\n","      <td>...</td>\n","      <td>-0.009431</td>\n","      <td>0.798278</td>\n","      <td>-0.137458</td>\n","      <td>0.141267</td>\n","      <td>-0.206010</td>\n","      <td>0.502292</td>\n","      <td>0.219422</td>\n","      <td>0.215153</td>\n","      <td>69.99</td>\n","      <td>0</td>\n","    </tr>\n","  </tbody>\n","</table>\n","<p>5 rows × 31 columns</p>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-64766cb3-5f58-4ab4-ae53-96500bb40306')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-64766cb3-5f58-4ab4-ae53-96500bb40306 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-64766cb3-5f58-4ab4-ae53-96500bb40306');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":4}]},{"cell_type":"markdown","source":["* Due to security issues, the original features V1, V2, ... V28 have been modified by PCA. However, we can guess that these features could be credit card number, expiration date, CVV, cardholder name, transaction location, transaction date-time, etc.\n","* The only two features that have not been converted with PCA are `Time` and `Amount`. Therefore, we only need to focus on processing these two features"],"metadata":{"id":"yfZ_xcxhtRDO"}},{"cell_type":"markdown","source":["##Exploratory Data Analysis"],"metadata":{"id":"nrb2vLNQvnXm"}},{"cell_type":"code","source":["df.shape"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"N7i_AtjKvpoq","outputId":"69b1a991-75bd-4eb3-9b5c-04a3fb79f6e4","executionInfo":{"status":"ok","timestamp":1659223922684,"user_tz":-420,"elapsed":28,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":5,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(284807, 31)"]},"metadata":{},"execution_count":5}]},{"cell_type":"code","source":["df.info()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"4XFVzZLcvqTp","outputId":"67b37085-7b9a-4042-9e9e-71668f53ef2a","executionInfo":{"status":"ok","timestamp":1659223922684,"user_tz":-420,"elapsed":23,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":6,"outputs":[{"output_type":"stream","name":"stdout","text":["<class 'pandas.core.frame.DataFrame'>\n","RangeIndex: 284807 entries, 0 to 284806\n","Data columns (total 31 columns):\n"," #   Column  Non-Null Count   Dtype  \n","---  ------  --------------   -----  \n"," 0   time    284807 non-null  float64\n"," 1   v1      284807 non-null  float64\n"," 2   v2      284807 non-null  float64\n"," 3   v3      284807 non-null  float64\n"," 4   v4      284807 non-null  float64\n"," 5   v5      284807 non-null  float64\n"," 6   v6      284807 non-null  float64\n"," 7   v7      284807 non-null  float64\n"," 8   v8      284807 non-null  float64\n"," 9   v9      284807 non-null  float64\n"," 10  v10     284807 non-null  float64\n"," 11  v11     284807 non-null  float64\n"," 12  v12     284807 non-null  float64\n"," 13  v13     284807 non-null  float64\n"," 14  v14     284807 non-null  float64\n"," 15  v15     284807 non-null  float64\n"," 16  v16     284807 non-null  float64\n"," 17  v17     284807 non-null  float64\n"," 18  v18     284807 non-null  float64\n"," 19  v19     284807 non-null  float64\n"," 20  v20     284807 non-null  float64\n"," 21  v21     284807 non-null  float64\n"," 22  v22     284807 non-null  float64\n"," 23  v23     284807 non-null  float64\n"," 24  v24     284807 non-null  float64\n"," 25  v25     284807 non-null  float64\n"," 26  v26     284807 non-null  float64\n"," 27  v27     284807 non-null  float64\n"," 28  v28     284807 non-null  float64\n"," 29  amount  284807 non-null  float64\n"," 30  class   284807 non-null  int64  \n","dtypes: float64(30), int64(1)\n","memory usage: 67.4 MB\n"]}]},{"cell_type":"code","source":["df.isnull().sum().sum()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"cM2Wn7e5vrW5","outputId":"fe058302-7531-4c3a-bccc-bff5ca2c32b4","executionInfo":{"status":"ok","timestamp":1659223922684,"user_tz":-420,"elapsed":17,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":7,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0"]},"metadata":{},"execution_count":7}]},{"cell_type":"markdown","source":["This dataset has 29 `float` variables, 1 target variable is `int` and no missing values"],"metadata":{"id":"s8xsoW1_v1yS"}},{"cell_type":"code","source":["df['class'].value_counts()/len(df)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"X52cpaNVvzl8","outputId":"a1c29d87-5c5a-40ab-b4dd-fbc1738b9d04","executionInfo":{"status":"ok","timestamp":1659223922685,"user_tz":-420,"elapsed":14,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":8,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0    0.998273\n","1    0.001727\n","Name: class, dtype: float64"]},"metadata":{},"execution_count":8}]},{"cell_type":"code","source":["f,ax = plt.subplots(figsize=(6,8))\n","ax = sns.countplot(x='class', data=df, palette='Set1')\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":497},"id":"ChT4C3OnwbQe","outputId":"29989a39-76cc-475c-9a9a-5443d36cec15","executionInfo":{"status":"ok","timestamp":1659223922685,"user_tz":-420,"elapsed":10,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":9,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 432x576 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["* This dataset contains transaction information with 492 TH frauds out of a total of 284,807 transactions. This is a severely imbalanced dataset with the positive class (fraud) accounting for 0.172% of all transactions"],"metadata":{"id":"prGTQY5Mwxl3"}},{"cell_type":"code","source":["df[['time','amount']].describe()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":300},"id":"ppnqo05iwe3W","outputId":"2708f020-5301-4148-c238-3f1dc096dfc0","executionInfo":{"status":"ok","timestamp":1659223923170,"user_tz":-420,"elapsed":494,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":10,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                time         amount\n","count  284807.000000  284807.000000\n","mean    94813.859575      88.349619\n","std     47488.145955     250.120109\n","min         0.000000       0.000000\n","25%     54201.500000       5.600000\n","50%     84692.000000      22.000000\n","75%    139320.500000      77.165000\n","max    172792.000000   25691.160000"],"text/html":["\n","  <div id=\"df-f3553992-0464-4978-9502-959dd9ec24bd\">\n","    <div class=\"colab-df-container\">\n","      <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>time</th>\n","      <th>amount</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>count</th>\n","      <td>284807.000000</td>\n","      <td>284807.000000</td>\n","    </tr>\n","    <tr>\n","      <th>mean</th>\n","      <td>94813.859575</td>\n","      <td>88.349619</td>\n","    </tr>\n","    <tr>\n","      <th>std</th>\n","      <td>47488.145955</td>\n","      <td>250.120109</td>\n","    </tr>\n","    <tr>\n","      <th>min</th>\n","      <td>0.000000</td>\n","      <td>0.000000</td>\n","    </tr>\n","    <tr>\n","      <th>25%</th>\n","      <td>54201.500000</td>\n","      <td>5.600000</td>\n","    </tr>\n","    <tr>\n","      <th>50%</th>\n","      <td>84692.000000</td>\n","      <td>22.000000</td>\n","    </tr>\n","    <tr>\n","      <th>75%</th>\n","      <td>139320.500000</td>\n","      <td>77.165000</td>\n","    </tr>\n","    <tr>\n","      <th>max</th>\n","      <td>172792.000000</td>\n","      <td>25691.160000</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-f3553992-0464-4978-9502-959dd9ec24bd')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-f3553992-0464-4978-9502-959dd9ec24bd button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-f3553992-0464-4978-9502-959dd9ec24bd');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":10}]},{"cell_type":"markdown","source":["**Observations:**\n","* On average, each transaction happens every 94813s\n","* The TB transaction amount is 88.34 with a standard deviation of 250, the maximum transaction amount is 25691\n","* Based on mean and max, it is easy to see that these two variables have outliers"],"metadata":{"id":"1ypH1Xr20cwX"}},{"cell_type":"code","source":["fraud = df[df['class']==1]\n","normal = df[df['class']==0]\n","print(f'Shape of Fraud Transactions: {fraud.shape}')\n","print(f'Shape of Normal Transactions: {normal.shape}')"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"I15RqJ1Y3N_g","outputId":"396ea21d-a727-4835-c946-d0d5f175b348","executionInfo":{"status":"ok","timestamp":1659223923171,"user_tz":-420,"elapsed":16,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":11,"outputs":[{"output_type":"stream","name":"stdout","text":["Shape of Fraud Transactions: (492, 31)\n","Shape of Normal Transactions: (284315, 31)\n"]}]},{"cell_type":"code","source":["pd.concat([fraud['amount'].describe(), normal['amount'].describe()], axis=1)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":300},"id":"-h1n9Vl_30dO","outputId":"27fca727-f32d-417f-8540-d2c7f7152aad","executionInfo":{"status":"ok","timestamp":1659223923171,"user_tz":-420,"elapsed":12,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":12,"outputs":[{"output_type":"execute_result","data":{"text/plain":["            amount         amount\n","count   492.000000  284315.000000\n","mean    122.211321      88.291022\n","std     256.683288     250.105092\n","min       0.000000       0.000000\n","25%       1.000000       5.650000\n","50%       9.250000      22.000000\n","75%     105.890000      77.050000\n","max    2125.870000   25691.160000"],"text/html":["\n","  <div id=\"df-e2929bc5-c3c1-4085-a019-2c263bff2f1f\">\n","    <div class=\"colab-df-container\">\n","      <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>amount</th>\n","      <th>amount</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>count</th>\n","      <td>492.000000</td>\n","      <td>284315.000000</td>\n","    </tr>\n","    <tr>\n","      <th>mean</th>\n","      <td>122.211321</td>\n","      <td>88.291022</td>\n","    </tr>\n","    <tr>\n","      <th>std</th>\n","      <td>256.683288</td>\n","      <td>250.105092</td>\n","    </tr>\n","    <tr>\n","      <th>min</th>\n","      <td>0.000000</td>\n","      <td>0.000000</td>\n","    </tr>\n","    <tr>\n","      <th>25%</th>\n","      <td>1.000000</td>\n","      <td>5.650000</td>\n","    </tr>\n","    <tr>\n","      <th>50%</th>\n","      <td>9.250000</td>\n","      <td>22.000000</td>\n","    </tr>\n","    <tr>\n","      <th>75%</th>\n","      <td>105.890000</td>\n","      <td>77.050000</td>\n","    </tr>\n","    <tr>\n","      <th>max</th>\n","      <td>2125.870000</td>\n","      <td>25691.160000</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-e2929bc5-c3c1-4085-a019-2c263bff2f1f')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-e2929bc5-c3c1-4085-a019-2c263bff2f1f button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-e2929bc5-c3c1-4085-a019-2c263bff2f1f');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":12}]},{"cell_type":"code","source":["fraud['amount'].value_counts().head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"anuGv13x5dJu","outputId":"087221e3-49ed-4014-abfd-b76012121939","executionInfo":{"status":"ok","timestamp":1659223923172,"user_tz":-420,"elapsed":11,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":13,"outputs":[{"output_type":"execute_result","data":{"text/plain":["1.00     113\n","0.00      27\n","99.99     27\n","0.76      17\n","0.77      10\n","Name: amount, dtype: int64"]},"metadata":{},"execution_count":13}]},{"cell_type":"markdown","source":["**Observations:**\n","* With fraud transactions, the average amount of fraud is 122.22 USD, the highest is 2125 USD, the lowest is 0 and the maximum amount is 1 USD with 113 times.\n"],"metadata":{"id":"aKcKU_Ei5UGs"}},{"cell_type":"markdown","source":["###Distribution of Time"],"metadata":{"id":"Cq7WoanD9qQl"}},{"cell_type":"code","source":["plt.figure(figsize=(8,6))\n","plt.title('Distribution of Transaction Time', fontsize=14)\n","sns.distplot(df['time'], bins=100)\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":405},"id":"-0fBwKPg9tie","outputId":"882bba5b-1904-4acc-d874-881a9c59a583","executionInfo":{"status":"ok","timestamp":1659223927446,"user_tz":-420,"elapsed":4281,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":14,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 576x432 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["**Observations:**\n","* This data set contains two-day trading information, looking at the distribution chart we see two peaks and two troughs. Most likely, the two peaks are transactions during the day because of the high volume of transactions, and the two bottoms are transactions at night when everyone is asleep."],"metadata":{"id":"wWA__IG-9yPJ"}},{"cell_type":"code","source":["fig, axs = plt.subplots(ncols=2, figsize=(16,4))\n","\n","sns.distplot(fraud['time'], bins=100, color='red', ax=axs[0])\n","axs[0].set_title(\"Distribution of Fraud Transactions\")\n","\n","sns.distplot(normal['time'], bins=100, color='green', ax=axs[1])\n","axs[1].set_title(\"Distribution of Genuine Transactions\")\n","\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":295},"id":"T9XL6mqNAxar","outputId":"9ff66949-bfeb-4089-f53f-7f5324fbe9a3","executionInfo":{"status":"ok","timestamp":1659223927447,"user_tz":-420,"elapsed":7,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":15,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1152x288 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["###Distribution of Amount"],"metadata":{"id":"g1wjtXiyF3wr"}},{"cell_type":"code","source":["fig, axs = plt.subplots(ncols=2,figsize=(16,4))\n","sns.distplot(fraud['amount'], bins=100, ax=axs[0])\n","axs[0].set_title(\"Distribution of Fraud Transactions\")\n","\n","sns.distplot(normal['amount'], bins=100, ax=axs[1])\n","axs[1].set_title(\"Distribution of Normal Transactions\")\n","\n","plt.show()"],"metadata":{"id":"1xXdoklwF6dn","colab":{"base_uri":"https://localhost:8080/","height":295},"outputId":"8466de1d-2b48-47dd-8018-b6feff0a6269","executionInfo":{"status":"ok","timestamp":1659223929822,"user_tz":-420,"elapsed":2380,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":16,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1152x288 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":["# Scale amount by log\n","df['amount_log'] = np.log(df.amount + 0.0001)"],"metadata":{"id":"NCdGANyL77qz","executionInfo":{"status":"ok","timestamp":1659223929823,"user_tz":-420,"elapsed":21,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":17,"outputs":[]},{"cell_type":"markdown","source":["Log transforms are useful when applied to skewed distributions because they tend to expand values in the lower magnitude range and tend to compress or reduce values in the magnitude range. higher"],"metadata":{"id":"4R0FPVR57zbc"}},{"cell_type":"code","source":["df.head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":299},"id":"2GQ13FkfBSV5","outputId":"4ea4c6ab-1e6c-4e1f-f251-9f5549790861","executionInfo":{"status":"ok","timestamp":1659223929823,"user_tz":-420,"elapsed":19,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":18,"outputs":[{"output_type":"execute_result","data":{"text/plain":["   time        v1        v2        v3        v4        v5        v6        v7  \\\n","0   0.0 -1.359807 -0.072781  2.536347  1.378155 -0.338321  0.462388  0.239599   \n","1   0.0  1.191857  0.266151  0.166480  0.448154  0.060018 -0.082361 -0.078803   \n","2   1.0 -1.358354 -1.340163  1.773209  0.379780 -0.503198  1.800499  0.791461   \n","3   1.0 -0.966272 -0.185226  1.792993 -0.863291 -0.010309  1.247203  0.237609   \n","4   2.0 -1.158233  0.877737  1.548718  0.403034 -0.407193  0.095921  0.592941   \n","\n","         v8        v9  ...       v22       v23       v24       v25       v26  \\\n","0  0.098698  0.363787  ...  0.277838 -0.110474  0.066928  0.128539 -0.189115   \n","1  0.085102 -0.255425  ... -0.638672  0.101288 -0.339846  0.167170  0.125895   \n","2  0.247676 -1.514654  ...  0.771679  0.909412 -0.689281 -0.327642 -0.139097   \n","3  0.377436 -1.387024  ...  0.005274 -0.190321 -1.175575  0.647376 -0.221929   \n","4 -0.270533  0.817739  ...  0.798278 -0.137458  0.141267 -0.206010  0.502292   \n","\n","        v27       v28  amount  class  amount_log  \n","0  0.133558 -0.021053  149.62      0    5.008099  \n","1 -0.008983  0.014724    2.69      0    0.989578  \n","2 -0.055353 -0.059752  378.66      0    5.936639  \n","3  0.062723  0.061458  123.50      0    4.816242  \n","4  0.219422  0.215153   69.99      0    4.248354  \n","\n","[5 rows x 32 columns]"],"text/html":["\n","  <div id=\"df-d9b298f6-d540-4651-b898-ea024902c372\">\n","    <div class=\"colab-df-container\">\n","      <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>time</th>\n","      <th>v1</th>\n","      <th>v2</th>\n","      <th>v3</th>\n","      <th>v4</th>\n","      <th>v5</th>\n","      <th>v6</th>\n","      <th>v7</th>\n","      <th>v8</th>\n","      <th>v9</th>\n","      <th>...</th>\n","      <th>v22</th>\n","      <th>v23</th>\n","      <th>v24</th>\n","      <th>v25</th>\n","      <th>v26</th>\n","      <th>v27</th>\n","      <th>v28</th>\n","      <th>amount</th>\n","      <th>class</th>\n","      <th>amount_log</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>0.0</td>\n","      <td>-1.359807</td>\n","      <td>-0.072781</td>\n","      <td>2.536347</td>\n","      <td>1.378155</td>\n","      <td>-0.338321</td>\n","      <td>0.462388</td>\n","      <td>0.239599</td>\n","      <td>0.098698</td>\n","      <td>0.363787</td>\n","      <td>...</td>\n","      <td>0.277838</td>\n","      <td>-0.110474</td>\n","      <td>0.066928</td>\n","      <td>0.128539</td>\n","      <td>-0.189115</td>\n","      <td>0.133558</td>\n","      <td>-0.021053</td>\n","      <td>149.62</td>\n","      <td>0</td>\n","      <td>5.008099</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>0.0</td>\n","      <td>1.191857</td>\n","      <td>0.266151</td>\n","      <td>0.166480</td>\n","      <td>0.448154</td>\n","      <td>0.060018</td>\n","      <td>-0.082361</td>\n","      <td>-0.078803</td>\n","      <td>0.085102</td>\n","      <td>-0.255425</td>\n","      <td>...</td>\n","      <td>-0.638672</td>\n","      <td>0.101288</td>\n","      <td>-0.339846</td>\n","      <td>0.167170</td>\n","      <td>0.125895</td>\n","      <td>-0.008983</td>\n","      <td>0.014724</td>\n","      <td>2.69</td>\n","      <td>0</td>\n","      <td>0.989578</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>1.0</td>\n","      <td>-1.358354</td>\n","      <td>-1.340163</td>\n","      <td>1.773209</td>\n","      <td>0.379780</td>\n","      <td>-0.503198</td>\n","      <td>1.800499</td>\n","      <td>0.791461</td>\n","      <td>0.247676</td>\n","      <td>-1.514654</td>\n","      <td>...</td>\n","      <td>0.771679</td>\n","      <td>0.909412</td>\n","      <td>-0.689281</td>\n","      <td>-0.327642</td>\n","      <td>-0.139097</td>\n","      <td>-0.055353</td>\n","      <td>-0.059752</td>\n","      <td>378.66</td>\n","      <td>0</td>\n","      <td>5.936639</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>1.0</td>\n","      <td>-0.966272</td>\n","      <td>-0.185226</td>\n","      <td>1.792993</td>\n","      <td>-0.863291</td>\n","      <td>-0.010309</td>\n","      <td>1.247203</td>\n","      <td>0.237609</td>\n","      <td>0.377436</td>\n","      <td>-1.387024</td>\n","      <td>...</td>\n","      <td>0.005274</td>\n","      <td>-0.190321</td>\n","      <td>-1.175575</td>\n","      <td>0.647376</td>\n","      <td>-0.221929</td>\n","      <td>0.062723</td>\n","      <td>0.061458</td>\n","      <td>123.50</td>\n","      <td>0</td>\n","      <td>4.816242</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>2.0</td>\n","      <td>-1.158233</td>\n","      <td>0.877737</td>\n","      <td>1.548718</td>\n","      <td>0.403034</td>\n","      <td>-0.407193</td>\n","      <td>0.095921</td>\n","      <td>0.592941</td>\n","      <td>-0.270533</td>\n","      <td>0.817739</td>\n","      <td>...</td>\n","      <td>0.798278</td>\n","      <td>-0.137458</td>\n","      <td>0.141267</td>\n","      <td>-0.206010</td>\n","      <td>0.502292</td>\n","      <td>0.219422</td>\n","      <td>0.215153</td>\n","      <td>69.99</td>\n","      <td>0</td>\n","      <td>4.248354</td>\n","    </tr>\n","  </tbody>\n","</table>\n","<p>5 rows × 32 columns</p>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d9b298f6-d540-4651-b898-ea024902c372')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-d9b298f6-d540-4651-b898-ea024902c372 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-d9b298f6-d540-4651-b898-ea024902c372');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":18}]},{"cell_type":"code","source":["# Correlation matrix \n","f, (ax1, ax2) = plt.subplots(1,2,figsize =( 18, 8))\n","corr = df.corr()\n","mask = np.zeros_like(corr, dtype=np.bool)\n","mask[np.triu_indices_from(mask)] = True\n","sns.heatmap((df.loc[df['class'] ==1]).corr(), vmax = .8, square=True, ax = ax1, cmap = 'afmhot', mask=mask);\n","ax1.set_title('Fraud')\n","sns.heatmap((df.loc[df['class'] ==0]).corr(), vmax = .8, square=True, ax = ax2, cmap = 'YlGnBu', mask=mask);\n","ax2.set_title('Normal')\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":501},"id":"boeqhTBy5KCk","executionInfo":{"status":"ok","timestamp":1659231360347,"user_tz":-420,"elapsed":4956,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}},"outputId":"98dbaac9-b070-4556-e037-1c533b94a035"},"execution_count":59,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1296x576 with 4 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["##Spliting the data"],"metadata":{"id":"99gjO7BpBirD"}},{"cell_type":"code","source":["X = df.drop(['class','time'], axis=1)\n","\n","y = df['class']"],"metadata":{"id":"CtjrENUYBiRD","executionInfo":{"status":"ok","timestamp":1659223929824,"user_tz":-420,"elapsed":16,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":19,"outputs":[]},{"cell_type":"code","source":["# split X and y into training and testing sets\n","\n","from sklearn.model_selection import train_test_split\n","\n","X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)\n","\n","X_train.shape, X_test.shape"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"cnC683hoBxdP","outputId":"9416051f-5277-4393-ef5f-da74bc8a05e2","executionInfo":{"status":"ok","timestamp":1659223929825,"user_tz":-420,"elapsed":16,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":20,"outputs":[{"output_type":"execute_result","data":{"text/plain":["((227845, 30), (56962, 30))"]},"metadata":{},"execution_count":20}]},{"cell_type":"markdown","source":["##Scaling the data"],"metadata":{"id":"o3MISRE-B9-X"}},{"cell_type":"code","source":["from sklearn.preprocessing import RobustScaler\n","\n","scaler = RobustScaler()\n","\n","X_train = scaler.fit_transform(X_train)\n","\n","X_test = scaler.transform(X_test)"],"metadata":{"id":"7pafbag8B7tf","executionInfo":{"status":"ok","timestamp":1659223930294,"user_tz":-420,"elapsed":479,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":21,"outputs":[]},{"cell_type":"markdown","source":["##SMOTE"],"metadata":{"id":"txH-jqFICE84"}},{"cell_type":"code","source":["from collections import Counter\n","from imblearn.over_sampling import SMOTE\n","print('Original dataset shape %s' % Counter(y_train))\n","\n","smt = SMOTE(random_state=42)\n","X_train_smt, y_train_smt = smt.fit_resample(X_train, y_train)\n","\n","print('Resampled dataset shape %s' % Counter(y_train_smt))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"_z5JzL0kCGvr","outputId":"1cc303f0-cf05-46e1-e280-2c345f4a7af8","executionInfo":{"status":"ok","timestamp":1659223930755,"user_tz":-420,"elapsed":467,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":22,"outputs":[{"output_type":"stream","name":"stdout","text":["Original dataset shape Counter({0: 227454, 1: 391})\n","Resampled dataset shape Counter({0: 227454, 1: 227454})\n"]}]},{"cell_type":"markdown","source":["##Modeling"],"metadata":{"id":"a7aTGku3Dr3t"}},{"cell_type":"code","source":["def plot_roc_cur(fper, tper):  \n","    plt.plot(fper, tper, color='orange', label='ROC')\n","    plt.plot([0, 1], [0, 1], color='darkblue', linestyle='--')\n","    plt.xlabel('False Positive Rate')\n","    plt.ylabel('True Positive Rate')\n","    plt.title('Receiver Operating Characteristic (ROC) Curve')\n","    plt.legend()\n","    plt.show()"],"metadata":{"id":"yAQbzTfUDtew","executionInfo":{"status":"ok","timestamp":1659223930756,"user_tz":-420,"elapsed":3,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":23,"outputs":[]},{"cell_type":"code","source":["import time\n","from sklearn.metrics import accuracy_score, roc_auc_score, precision_score, recall_score, f1_score, cohen_kappa_score, plot_confusion_matrix, roc_curve, classification_report\n","def run_model(model, X_train, y_train, X_test, y_test, verbose=True):\n","    t0=time.time()\n","    if verbose == False:\n","        model.fit(X_train,y_train, verbose=0)\n","    else:\n","        model.fit(X_train,y_train)\n","    y_pred = model.predict(X_test)\n","    accuracy = accuracy_score(y_test, y_pred)\n","    roc_auc = roc_auc_score(y_test, y_pred)\n","    precision = precision_score(y_test, y_pred)\n","    recall = recall_score(y_test, y_pred)\n","    f1 = f1_score(y_test, y_pred)\n","    coh_kap = cohen_kappa_score(y_test, y_pred)\n","    time_taken = time.time()-t0\n","    print(\"Accuracy = {}\".format(accuracy))\n","    print(\"ROC Area under Curve = {}\".format(roc_auc))\n","    #print(\"Precision  = {}\".format(precision))\n","    #print(\"Recall  = {}\".format(recall))\n","    print(\"F1 Score  = {}\".format(f1))\n","    print(\"Cohen's Kappa = {}\".format(coh_kap))\n","    print(\"Time taken = {}\".format(time_taken))\n","    print(classification_report(y_test,y_pred,digits=5))\n","    \n","    probs = model.predict_proba(X_test)  \n","    probs = probs[:, 1]  \n","    fper, tper, thresholds = roc_curve(y_test, probs) \n","    plot_roc_cur(fper, tper)\n","    \n","    plot_confusion_matrix(model, X_test, y_test,cmap=plt.cm.Blues)\n","    \n","    return model, accuracy, roc_auc, f1, coh_kap, time_taken"],"metadata":{"id":"dN0H_rm5DwmK","executionInfo":{"status":"ok","timestamp":1659231198077,"user_tz":-420,"elapsed":303,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":57,"outputs":[]},{"cell_type":"markdown","source":["###Logistic Regression"],"metadata":{"id":"itf8NUX_HZKm"}},{"cell_type":"code","source":["# Logistic Regression with SMOTE\n","from sklearn.linear_model import LogisticRegression\n","\n","params_lr = {'penalty': 'l1', 'solver':'liblinear'}\n","\n","model_lrsmt = LogisticRegression(**params_lr)\n","model_lrsmt, accuracy_lrsmt, roc_auc_lrsmt,f1_score_lrsmt,coh_kap_lrsmt, tt_lrsmt = run_model(model_lrsmt, X_train_smt, y_train_smt, X_test, y_test)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":800},"id":"KEMbXrFmHasZ","outputId":"21fc94c8-18e4-401d-9fab-bc700f2c877d","executionInfo":{"status":"ok","timestamp":1659223948966,"user_tz":-420,"elapsed":18212,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":25,"outputs":[{"output_type":"stream","name":"stdout","text":["Accuracy = 0.9757908781292791\n","ROC Area under Curve = 0.9582237281430259\n","F1 Score  = 0.12109623964308477\n","Cohen's Kappa = 0.11817037698754529\n","Time taken = 17.78870129585266\n","              precision    recall  f1-score   support\n","\n","           0    0.99989   0.97585   0.98773     56861\n","           1    0.06471   0.94059   0.12110       101\n","\n","    accuracy                        0.97579     56962\n","   macro avg    0.53230   0.95822   0.55441     56962\n","weighted avg    0.99823   0.97579   0.98619     56962\n","\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["###Decision Tree"],"metadata":{"id":"a9W4hnLeHnml"}},{"cell_type":"code","source":["# Decision Tree\n","from sklearn.tree import DecisionTreeClassifier\n","\n","params_dt = {'max_depth': 16,\n","             'max_features': \"sqrt\"}\n","\n","model_dtsmt = DecisionTreeClassifier(**params_dt)\n","model_dtsmt, accuracy_dtsmt, roc_auc_dtsmt, f1_score_dtsmt,coh_kap_dtsmt, tt_dtsmt = run_model(model_dtsmt, X_train_smt, y_train_smt, X_test, y_test)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":800},"id":"K3aAtl15HqeQ","outputId":"c0ba4093-a455-4888-88d5-e8e1bb0c0eb2","executionInfo":{"status":"ok","timestamp":1659223953651,"user_tz":-420,"elapsed":4694,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":26,"outputs":[{"output_type":"stream","name":"stdout","text":["Accuracy = 0.9902566623362944\n","ROC Area under Curve = 0.9012273459631712\n","F1 Score  = 0.22809457579972184\n","Cohen's Kappa = 0.225734556245497\n","Time taken = 4.400159120559692\n","              precision    recall  f1-score   support\n","\n","           0    0.99966   0.99057   0.99510     56861\n","           1    0.13269   0.81188   0.22809       101\n","\n","    accuracy                        0.99026     56962\n","   macro avg    0.56617   0.90123   0.61160     56962\n","weighted avg    0.99813   0.99026   0.99374     56962\n","\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["###Random Forest"],"metadata":{"id":"s-tlo-UBIBge"}},{"cell_type":"code","source":["# Random Forest with SMOTE\n","from sklearn.ensemble import RandomForestClassifier\n","\n","model_rfsmt = RandomForestClassifier()\n","model_rfsmt, accuracy_rfsmt, roc_auc_rfsmt, f1_score_rfsmt,coh_kap_rfsmt, tt_rfsmt = run_model(model_rfsmt, X_train_smt, y_train_smt, X_test, y_test)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":800},"id":"qy-cTFDCIpiq","outputId":"44b311a9-dc92-4dbc-a157-d2a4c82c9338","executionInfo":{"status":"ok","timestamp":1659224274076,"user_tz":-420,"elapsed":320429,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":27,"outputs":[{"output_type":"stream","name":"stdout","text":["Accuracy = 0.9995084442259752\n","ROC Area under Curve = 0.9206865587281543\n","F1 Score  = 0.8585858585858587\n","Cohen's Kappa = 0.8583397534497652\n","Time taken = 318.77797985076904\n","              precision    recall  f1-score   support\n","\n","           0    0.99972   0.99979   0.99975     56861\n","           1    0.87629   0.84158   0.85859       101\n","\n","    accuracy                        0.99951     56962\n","   macro avg    0.93800   0.92069   0.92917     56962\n","weighted avg    0.99950   0.99951   0.99950     56962\n","\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["###LightGBM"],"metadata":{"id":"9gufSfA7JEzR"}},{"cell_type":"code","source":["# Light GBM with SMOTE\n","import lightgbm as lgb\n","\n","model_lgbsmt = lgb.LGBMClassifier()\n","model_lgbsmt, accuracy_lgbsmt, roc_auc_lgbsmt, f1_score_lgbsmt,coh_kap_lgbsmt, tt_lgbsmt = run_model(model_lgbsmt, X_train_smt, y_train_smt, X_test, y_test)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":800},"id":"AsFwZElhJGkz","outputId":"a844755d-8a27-4f35-8329-8fcf3edb76f9","executionInfo":{"status":"ok","timestamp":1659224279528,"user_tz":-420,"elapsed":5461,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":28,"outputs":[{"output_type":"stream","name":"stdout","text":["Accuracy = 0.9985077771145676\n","ROC Area under Curve = 0.8954768280683083\n","F1 Score  = 0.6530612244897959\n","Cohen's Kappa = 0.6523365857658603\n","Time taken = 4.842337131500244\n","              precision    recall  f1-score   support\n","\n","           0    0.99963   0.99887   0.99925     56861\n","           1    0.55556   0.79208   0.65306       101\n","\n","    accuracy                        0.99851     56962\n","   macro avg    0.77759   0.89548   0.82616     56962\n","weighted avg    0.99884   0.99851   0.99864     56962\n","\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["###Catboost"],"metadata":{"id":"bxB-6pS6J76F"}},{"cell_type":"code","source":["!pip install catboost"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"gRNrUwbCKFbf","outputId":"8f29e173-6151-4e88-e1f4-744b878b00c6","executionInfo":{"status":"ok","timestamp":1659224289072,"user_tz":-420,"elapsed":9551,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":29,"outputs":[{"output_type":"stream","name":"stdout","text":["Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n","Collecting catboost\n","  Downloading catboost-1.0.6-cp37-none-manylinux1_x86_64.whl (76.6 MB)\n","\u001b[K     |████████████████████████████████| 76.6 MB 1.1 MB/s \n","\u001b[?25hRequirement already satisfied: numpy>=1.16.0 in /usr/local/lib/python3.7/dist-packages (from catboost) (1.21.6)\n","Requirement already satisfied: scipy in /usr/local/lib/python3.7/dist-packages (from catboost) (1.7.3)\n","Requirement already satisfied: plotly in /usr/local/lib/python3.7/dist-packages (from catboost) (5.5.0)\n","Requirement already satisfied: six in /usr/local/lib/python3.7/dist-packages (from catboost) (1.15.0)\n","Requirement already satisfied: matplotlib in /usr/local/lib/python3.7/dist-packages (from catboost) (3.2.2)\n","Requirement already satisfied: pandas>=0.24.0 in /usr/local/lib/python3.7/dist-packages (from catboost) (1.3.5)\n","Requirement already satisfied: graphviz in /usr/local/lib/python3.7/dist-packages (from catboost) (0.10.1)\n","Requirement already satisfied: pytz>=2017.3 in /usr/local/lib/python3.7/dist-packages (from pandas>=0.24.0->catboost) (2022.1)\n","Requirement already satisfied: python-dateutil>=2.7.3 in /usr/local/lib/python3.7/dist-packages (from pandas>=0.24.0->catboost) (2.8.2)\n","Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->catboost) (1.4.4)\n","Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.7/dist-packages (from matplotlib->catboost) (0.11.0)\n","Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->catboost) (3.0.9)\n","Requirement already satisfied: typing-extensions in /usr/local/lib/python3.7/dist-packages (from kiwisolver>=1.0.1->matplotlib->catboost) (4.1.1)\n","Requirement already satisfied: tenacity>=6.2.0 in /usr/local/lib/python3.7/dist-packages (from plotly->catboost) (8.0.1)\n","Installing collected packages: catboost\n","Successfully installed catboost-1.0.6\n"]}]},{"cell_type":"code","source":["# Catboost\n","import catboost as cb\n","params_cb ={'iterations': 20,\n","            'max_depth': 16}\n","\n","model_cbsmt = cb.CatBoostClassifier(**params_cb)\n","model_cbsmt, accuracy_cbsmt, roc_auc_cbsmt, f1_score_cbsmt,coh_kap_cbsmt, tt_cbsmt = run_model(model_cbsmt, X_train_smt, y_train_smt, X_test, y_test)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000},"id":"oQWyP7wbJ_Dn","outputId":"19e70783-a5ea-4f63-9d16-db9bd783517b","executionInfo":{"status":"ok","timestamp":1659224412675,"user_tz":-420,"elapsed":123619,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":30,"outputs":[{"output_type":"stream","name":"stdout","text":["Learning rate set to 0.5\n","0:\tlearn: 0.0939020\ttotal: 6.69s\tremaining: 2m 7s\n","1:\tlearn: 0.0369052\ttotal: 14.1s\tremaining: 2m 6s\n","2:\tlearn: 0.0199681\ttotal: 21.6s\tremaining: 2m 2s\n","3:\tlearn: 0.0125476\ttotal: 28.4s\tremaining: 1m 53s\n","4:\tlearn: 0.0086885\ttotal: 34s\tremaining: 1m 42s\n","5:\tlearn: 0.0059908\ttotal: 39.6s\tremaining: 1m 32s\n","6:\tlearn: 0.0042267\ttotal: 45.3s\tremaining: 1m 24s\n","7:\tlearn: 0.0034408\ttotal: 50.8s\tremaining: 1m 16s\n","8:\tlearn: 0.0028330\ttotal: 56.5s\tremaining: 1m 9s\n","9:\tlearn: 0.0023885\ttotal: 1m 2s\tremaining: 1m 2s\n","10:\tlearn: 0.0021404\ttotal: 1m 7s\tremaining: 55.3s\n","11:\tlearn: 0.0018040\ttotal: 1m 13s\tremaining: 48.8s\n","12:\tlearn: 0.0015751\ttotal: 1m 18s\tremaining: 42.4s\n","13:\tlearn: 0.0012856\ttotal: 1m 24s\tremaining: 36.2s\n","14:\tlearn: 0.0011690\ttotal: 1m 29s\tremaining: 30s\n","15:\tlearn: 0.0010272\ttotal: 1m 35s\tremaining: 23.9s\n","16:\tlearn: 0.0009326\ttotal: 1m 41s\tremaining: 17.8s\n","17:\tlearn: 0.0008497\ttotal: 1m 46s\tremaining: 11.9s\n","18:\tlearn: 0.0007541\ttotal: 1m 52s\tremaining: 5.92s\n","19:\tlearn: 0.0007540\ttotal: 1m 57s\tremaining: 0us\n","Accuracy = 0.9987711105649381\n","ROC Area under Curve = 0.9203172370489717\n","F1 Score  = 0.7083333333333333\n","Cohen's Kappa = 0.7077330595283873\n","Time taken = 122.86549234390259\n","              precision    recall  f1-score   support\n","\n","           0    0.99972   0.99905   0.99938     56861\n","           1    0.61151   0.84158   0.70833       101\n","\n","    accuracy                        0.99877     56962\n","   macro avg    0.80561   0.92032   0.85386     56962\n","weighted avg    0.99903   0.99877   0.99887     56962\n","\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["###XGBoost"],"metadata":{"id":"IPgk6px1KMZR"}},{"cell_type":"code","source":["# XGBoost with SMOTE\n","import xgboost as xgb\n","params_xgb ={'n_estimators': 20,\n","            'max_depth': 16}\n","\n","model_xgbsmt = xgb.XGBClassifier(**params_xgb)\n","model_xgbsmt, accuracy_xgbsmt, roc_auc_xgbsmt, f1_score_xgbsmt,coh_kap_xgbsmt, tt_xgbsmt = run_model(model_xgbsmt, X_train_smt, y_train_smt, X_test, y_test)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":800},"id":"2F2-MfVoKPDG","outputId":"47cf5862-30b3-47da-e806-d9d8677950e5","executionInfo":{"status":"ok","timestamp":1659224541563,"user_tz":-420,"elapsed":128899,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":31,"outputs":[{"output_type":"stream","name":"stdout","text":["Accuracy = 0.9984199992977775\n","ROC Area under Curve = 0.9102579662303122\n","F1 Score  = 0.6484375\n","Cohen's Kappa = 0.6476810261754515\n","Time taken = 127.98923969268799\n","              precision    recall  f1-score   support\n","\n","           0    0.99968   0.99873   0.99921     56861\n","           1    0.53548   0.82178   0.64844       101\n","\n","    accuracy                        0.99842     56962\n","   macro avg    0.76758   0.91026   0.82382     56962\n","weighted avg    0.99886   0.99842   0.99859     56962\n","\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd5jU5PbA8e9hKUvvFurSYSkCLiAgTaQpgl7LRbHgXUFERC+Wn4oFkYsVOyioFAEBxauioogF8YqF3qv0Xlx623J+fySrw7plgM1kZ+d8nmefnUwyyUkyk5P3fZM3oqoYY4yJXHn8DsAYY4y/LBEYY0yEs0RgjDERzhKBMcZEOEsExhgT4SwRGGNMhLNEkMOIyAoRaet3HDmFiDwqIu/4tOxxIjLUj2VnNxHpKSJfn+Vnz/o7KSI/iUijs/ns2RKRe0TkuVAuM9xZIsiEiGwSkeMickREdrkHhiJeLlNV66rqbC+XkUpECojIMyKyxV3PdSLyoIhIKJafTjxtRWRb4HuqOkxV7/BoeSIiA0RkuYgcFZFtIvKhiNT3YnlnS0QGi8jEc5mHqk5S1Y5BLOtvye9sv5MichVwWFUXucODRSTR/T0dEJG5ItI8zWdKiMib7u/tmIgsE5Hb05n3TSIy353XThH5UkQudUe/DfQUkfMyiS0s9n2oWCLI2lWqWgRoCDQCHvE5njMmInkzGPUh0B64AigK3AL0AV71IAYRkZz2fXsVuBcYAJQCagKfAFdm94Iy2Qee83HZfYEJad6b6v6eygDf43wHARCR/MA3QGWgOVAceBB4VkQGBkw3EHgFGAacD1QCRgLdAVT1BPAlcGsmsWXbvvdz32YbVbW/DP6ATcDlAcPPA18EDF8CzAUOAEuAtgHjSgFjgR1AAvBJwLiuwGL3c3OBBmmXCZQDjgOlAsY1AvYB+dzhfwGr3PnPBCoHTKvA3cA6YGM669YeOAFUTPN+MyAZqO4OzwaeAX4DDgGfpokps20wG/gP8JO7LtWB292YDwMbgDvdaQu706QAR9y/csBgYKI7TYy7XrcBW9xtMShgeQWB8e72WAU8BGzLYN/WcNezaSb7fxwwAvjCjfdXoFrA+FeBre52WQC0Chg3GJgGTHTH3wE0BX52t9VO4A0gf8Bn6gKzgD+A3cCjQGfgFJDobpMl7rTFgXfd+WwHhgJR7rhe7jZ/GdjvjusF/M8dL+64PW5sy4B6OCcBie7yjgCfpf0dAFFuXL+722QBab5D7nT53f1ZIc02mRgwHOvuz7LucLwbU+E08/qnG08xd72PANdn8dvtCXx/Dvt+NnBHwPCf2y+93xfwJvBimnl8Cgx0X5cDPgL2utMP8Pv4dlqsfgeQk//S/AAquD+YV93h8u6P7AqcklUHdzj1S/0FMBUoCeQD2rjvN3K/7M3cH9Vt7nIKpLPM74DeAfG8ALzlvu4OrAfqAHmBx4C5ab6os3ASUsF01u1Z4IcM1nszfx2gZ+McaOrhHKw/4q8Dc1bbYDbOAbuuG2M+nDOuajgHozbAMaCxO31b0hy4ST8RvI1z0L8IOAnUCVwnd5tXAJamnV/AfPsCm7PY/+Pc9Wnqxj8JmBIw/magtDvufmAXEB0QdyJwtbttCgIX4yTOvO66rALuc6cvinNQvx+Idoebpd0GAcv+GBjl7pPzcBJ16j7rBSQB97jLKsjpiaATzgG8hLsf6gAXBqzz0Ex+Bw/i/A5quZ+9CCidzrarCxzNZF/md/fXPiCv+94UYHw688rrrk8nnMSYlPqZTPZdY+CPc9j3s8k6Efz5+wJa45wUiDu+JE4iLOfu/wXAE+56V8U5Cerk9zEu9S+nFdVzok9E5DDOTt4DPOm+fzMwQ1VnqGqKqs4C5gNXiMiFQBegr6omqGqiqv7gfq4PMEpVf1XVZFUdj3MwuySdZb8P3AhO1QrQw30PnC/zM6q6SlWTcIrJDUWkcsDnn1HVP1T1eDrzLoNz4EnPTnd8qgmqulxVjwKPAzeISFRm2yDgs+NUdYWqJrnb4QtV/V0dPwBfA60yiCMjT6nqcVVdglMKuch9/wZgmLvNtwGvZTKP0pmsf6CPVfU3dxtPwqkiBEBVJ6rqfnfdhgMFcA6QqX5W1U/cbXNcVReo6i/u9JtwDuRt3Gm7ArtUdbiqnlDVw6r6a3oBicj5ONv4PlU9qqp7cM7wewRMtkNVX3eXlXb/J+Ikmto4B65VqhrMtgCnZPOYqq5x9+ESVd2fznQlcEoMad0gIgdwDpK9gevcbQsZfCfd8fvc8aWBfQGfychhnNJDeoLd91kJ/H39iJMcUr/L1+Hs/x1AE5yToyGqekpVN+CczPRId64+sESQtatVtSjO2Wpt/jpAVgaudxu9Drhf7kuBC4GKOGcjCenMrzJwf5rPVcQ5c0jrI6C5m1ha41Sb/Bgwn1cD5vEHzhla+YDPb81kvfa5sabnQnd8evPZjHNmX4bMt0G6MYhIFxH5RUT+cKe/gtOTTjB2Bbw+BqQ24JdLs7zM1n8/Ga9/MMtCRB4QkVUictBdl+Kcvi5p172miHzuNoQewkneqdNXxKluCUZlnH2wM2C7j8IpGaS77ECq+h1OtdQIYI+IjBaRYkEuO9g4E3CSTVofqGoJnLr95TilpFTpfifdOvgy7vj9QJkg6uWLAgczGBfsvs/Kn9tYnWLAFNwTN+AmnBMHcPZXuTS/k0dxtkGOYIkgSO7Z6zjgRfetrThnyiUC/gqr6rPuuFIiUiKdWW0F/pPmc4VUdXI6y0zAOWP+J84Xa4r7hUudz51p5lNQVecGziKTVfoGaCYiFQPfFJFmOD/27wLeDpymEs4Z5b4stsHfYhCRAjjJ7UXgfPeAMAMngWUVbzB24lQJpRd3Wt8CFUQk7mwWJCKtcNogbgBKuutykL/WBf6+Pm8Cq4EaqloM52CQOv1WnCqD9KSdz1acUmSZgO1eTFXrZvKZ02eo+pqqXoxTT18Tp8ony8+5y66WxTTgVFuKiJRPb6Sq7sMpHQ92T3TA+U52EZHCaSa/Fmd9f8FpYzmJU+WWmTo4pcX0BLPvjwKFAoYvSGeatNtqMnCdWypvhvNdB2ebbUzzOymqqleQQ1giODOvAB1E5CKcRsCrRKSTiESJSLR7+WMFt5j9JTBSREqKSD4Rae3O422gr4g0c6+kKSwiV4pIemdP4FQF3YpT1Hw/4P23gEdEpC6AiBQXkeuDXRFV/QbnB/GRiNR11+ESd73eVNV1AZPfLCKxIlIIGAJMU9XkzLZBBovNj1N9shdIEpEuQOAljbuB0iKSUZE+Kx/gbJOS7gGof0YTuus3Epjsxpzfjb+HiDwcxLKK4tRV7wXyisgTOI2ZWX3mEHBERGoDdwWM+xy4UETuE+ey3qJuUgZnu8SkXnXlfr++BoaLSDERySMi1USkDUEQkSbu9y8fzgHvBE5pM3VZGSUkgHeAp0Wkhvv9bSAipdNOpKqncA7sGcakqmtwLnJ4yH1rArAN+FBEYtzfTSecKr7BqnpQVQ/i1LWPEJGrRaSQO10XEXk+YPZtcH6D6S03mH2/GPiHO//qOA3ZmVLnMtl97jaaqaoH3FG/AYdF5P9EpKD7W6knIk2ymmeoWCI4A6q6F3gPeEJVt+I02D6KczDYinNWlbpNb8E5c16N07ZwnzuP+Th1o2/gFJ/X4zREZWQ6zlUOu9w68dRYPgaeA6a41QzLcdolzsS1OJfwfYVzJcZEnCtR7kkz3QSc0tAunIbMAW4MWW2D06jqYfezH+Cs+03u+qWOX41zVrXBLUKnV12WmSE4B5KNOAehaThnjxkZwF9VJAdwqjyuAT4LYlkzcbbbWpzqshNkXhUF8ADOOh/GOSGYmjrC3TYdgKtwtvM6oJ07OvUSy/0istB9fStOYl2Jsy2nEXx1RzF3+Qlu7PtxLkQAZ//Hutv/k3Q++xLO/vsaJ6m9i9NYmp5ROL+DzLwA9BGR81T1JM4Vc1txrtA65C5vkKqmxofbHjMQ5wKJ1O9df5zLPxGRaJwqx/GZLDerff8yztVTu935TEpnHul5312HP0/a3JOmrjjtSxv5K1mc7QlPtktt4TYmXSIyG+dKD1/u7j0XInIX0ENVgzpTNtlPRH4C+rtny6Fa5j04l7Q+lOXEBnAuyzImV3Drmqvi1CPXwLkU8w1fg4pwqtrSh2W+HuplhjtLBCY3yY9THVEFp7g/Bacu2BiTCasaMsaYCGeNxcYYE+HCrmqoTJkyGhMT43cYxhgTVhYsWLBPVcumNy7sEkFMTAzz58/3OwxjjAkrIrI5o3FWNWSMMRHOEoExxkQ4SwTGGBPhLBEYY0yEs0RgjDERzrNEICJjRGSPiCzPYLyIyGsisl5ElopIY69iMcYYkzEvSwTjcB4rl5EuOP3B1MDpl/xND2MxxhiTAc/uI1DVOSISk8kk3YH33Aet/CIiJUTkwjN4ZJ4xxuQ+pw7CupGQ/NcTRo8eg70JQkzTrlA6+x9j4OcNZeU5vf/2be57f0sEItIHp9RApUqVQhKcMcb4YtcsWPKoOyB8t6Iavd+5juKFTjD/q/nkyWWJIGiqOhoYDRAXF2e95Bljcp+EpbDxPTi0CoADLZfw4NCdvPPOMqpXL8HL73QiT63Mnr569vxMBNs5/ZmyFdz3jDEm8qx8BjZPhbyFSY6OoUXn31iz9iAPPdSEwYNbULBgPs8W7WcimA70F5EpOA96PmjtA8Z4LCUJTiW4f3/AyT/+ep06nJLZ0z2NZ3Z/x/4iV1Pqqo+IEuE/so6KFYsSF3eB54v2LBGIyGSgLVBGRLYBTwL5AFT1LWAGznNF1wPHgNu9isWYXEUVko6efgA/leAe1P/I+CB/KgESD2U+73zFICqjRxAbr6jCpDl1ufe9ljz7/DJ6927ANdfUCNnyvbxq6MYsxitwt1fLNybHS0mCUweyPninPcCf+gNSEjOeb558kL8U5C/p/C9YHorXgwKl3PfdcQVKnT5d/hKQJyyaDXOVrVsP0bfvLGbM2Mgll1xAy5blQh6D7XVjzoUqJB8L7ow87XAwZ+d/HqRLQfFyfz94Fyh1+jT5S0LewiASmvU352Ty5FXceecskpNTeOWVdvTv34ioqNB3+GCJIJLt+xVWvwwpp/yOJHykJEJiwukH9cy2n+Q9/eBd8EIoXjf9M/ICac/OvWscNDlDyZLRNGt2IaNHd6BKlRK+xWGJIFJtGA+/9XHOOgte6Hc04UOi3LPz2PTPyNMO29m5CZCUlMLLL8/n1KkUBg26hM6dq9CpUwzi83fEEkEobZwAmyb5HYVzx+KeOXB+e7j0A+fgZYzx1JIle4iPn8mCBbu54YZaqCoi4nsSAEsEobXxPdj1DZRu5nckEPsINHjKqh+M8djJk0kMHfoLzz77G6VKRfPhh1dx7bU1c0QCSGWJwCspyTC3JxzfBrg7/MAyKNMCOv7ka2jGmNBZty6B5577jZtuqs1LL7WjdOmcd3muJQKvnNwLW6Y6r8+/zPlf6mKodIN/MRljQuLIkVN8+ul6evaMpV69sqxe/S+qVvWvMTgrlgiyW+IhmHM1nNjrDDd5E2r09TcmY0zIzJq1iT59vmbz5kM0bnw+deqUztFJAOwJZdnv0FrY/T3kLQQVr3UaZI0xuV5Cwgni47+iY8dp5M8fxQ8/9KBOndJ+hxUUKxEE68AKmHsTJB0DySR/pvYhftF/4ILLQxObMcZXyckptGz5PmvXJvDII8144onmREeHz+E1fCL124ElcGAplGwMxWpmPm3eolC6aWjiMsb4Zt++Y5QqVZCoqDwMG9aKSpWK0bjx+X6HdcYsEaT6rqNzVY9EpT8++Zjzv+UUKBa6zqCMMTmPqjJhwkruu+97nn22FX36XMTVV4fvccESQapds5z/1eIznqZAGShSNTTxGGNypM2bD3LnnbOYOXMTLVqUo3XrCn6HdM4iNxEkLIXZnZ2rfMS9qarek9BgsK9hGWNyrokTV3LXXbNQhddfv4x+/RqRJ0/OuTHsbEVuIji8Fo7vdK7xL17PqRKqcrPfURljcrCyZQvSsmV5Ro3qQOXKxf0OJ9tEZiJY8rjzWDiAuNedDsSMMSaNxMRkhg+fT2JiCo8/3pxOnarQsaP/ncRlt8hMBAmLnJ4hYx+CYrX9jsYYkwMtWrSb+PiZLFq0hx49aueoTuKyW+TdULZ3Luz4AqLPgzoPZH5PgDEm4pw4kcSjj/5IkyYT2bHjCB991I3Jk7vmygSQKvJKBIfWOP9r3etvHMaYHGn9+gRefHEet95al+HD21KyZLTfIXkushLBhvHw67+c1xdY1w/GGMeRI6f4+ON13HJLXerVK8uaNf/y9YlhoRZZ9SJHNzn/m78Hhav4GooxJmeYOXMjdeuO5bbbvmTVqv0AEZUEINISQaqYm+3xgcZEuP37j3PbbTPo3PkjChXKx48/3hg2ncRlt8iqGlJ1/lsSMCaiOZ3ETWb9+gQGDbqExx67JKw6ictuEbbm6ncAxhgf7d17jNKlnU7innuuNZUrF6Nhw/P8Dst3kVU1lHzC7wiMMT5QVcaOXUbNmu/y9ttLAejevbolAVdklQgSFvkdgTEmxDZtOkifPl8za9ZmWrWqQLt2Ff0OKceJrESQJz/kKeB3FMaYEJkwYQV33fUNIjBy5OXceedFuaKTuOwWWYng0GqIKuh3FMaYEDn//MK0bl2Bt97qQKVKxfwOJ8eKnESgKXBsGxSv43ckxhiPJCYm8/zz80hOTuGJJ1rQsWMMHTvG+B1Wjhc5jcV7foCUk5DPzgqMyY0WLtxNkyYTeeyx/7FmTQKqdpVgsCInESQecf7XH+xrGMaY7HX8eCIPPzyHpk0nsnv3MT7+uDuTJl2ZqzuJy26eJgIR6Swia0RkvYg8nM74SiLyvYgsEpGlInKFl/EAViIwJpfZsOEgL700n1696rFy5e1h/exgv3iWCEQkChgBdAFigRtFJO0TYB4DPlDVRkAPYKRX8Rhjco9Dh04ybtxyAOrWLcO6dfG8806niOgp1AtelgiaAutVdYOqngKmAN3TTKNA6il6cWCHd+FYfaExucGMGRuoV28c8fEz/+wkLjc9NtIPXiaC8sDWgOFt7nuBBgM3i8g2YAZwT3ozEpE+IjJfRObv3bv3HMOyekNjwtG+fce45ZYZXHnlfylaND8//RS5ncRlN78bi28ExqlqBeAKYILI3x8ZpqqjVTVOVePKli0b8iCNMf5K7SRuypTVPPFEcxYuvIVLLinnd1i5hpf3EWwHAu/lruC+Fyge6Aygqj+LSDRQBtjjYVzGmDCxe/dRypYtRFRUHl58sS2VKxejQQM7GcxuXpYI5gE1RKSKiOTHaQyenmaaLUB7ABGpA0QD51r3kwFrIzAmXKgq7767jFq1xjB69BIArrqqmiUBj3hWIlDVJBHpD8wEooAxqrpCRIYA81V1OnA/8LaI/BvnSN1LPb8LxNoIjMnJNmw4QO/eX/Pdd1to06YCl19e2e+Qcj1Pu5hQ1Rk4jcCB7z0R8Hol0NLLGIwx4WP8+OX06/cNUVF5eOutDvTu3cA6iQuByOlryBiT45UrV4TLLqvEm292oEKFon6HEzEiJxFYvyPG5DinTiXz7LO/kpKiDB7ckg4dYujQIcbvsCKO35ePhp71P2JMjjBv3k4uvngCTz45lw0bDloncT6KvERgjPHVsWOJPPDAbC655H0SEk4wffo1vPfeFdZJnI8ip2rIGJMjbNx4kNdfX0Tv3g147rnWFC9uTw30WwQlAit2GuOXgwdP8t//ruX22+tTt24Z1q+Pp2JF6wk4p4jAqiErfhoTSl988Tt1647ljju+ZvVqp5M4SwI5SwQmAmNMKOzde4yePb+ga9ePKVkymp9/vonata2TuJwogqqGjDGhkpycwqWXTmbjxoM89VQLHn64GfnzR/kdlslABCUCayMwxmu7dh3lvPOcTuKGD29LTEwx6tWz/oFyuqCrhkSkkJeBhI61ERiT3VJSlFGjllCz5ruMGuV0Ete1azVLAmEiy0QgIi1EZCWw2h2+SETskZLGGADWr0+gffsP6Nt3Fk2aXECnTjF+h2TOUDBVQy8DnXC7kFbVJSLS2tOojDFhYezYZfTr9y358+fh7bc7Eh9f324MC0NBtRGo6tY0OzfZm3A8ZLevG5PtKlUqRqdOMYwY0Z7y5a2TuHAVTCLYKiItABWRfMC9wCpvw/KQna0Yc9ZOnkzimWecTuKGDLmU9u0r0769PS8g3AXTWNwXuBvnwfPbgYZAPy+DMsbkPL/+6nQS99RTP7Nly2HrJC4XCaZEUEtVewa+ISItgZ+8CckYk5McPXqKxx//iVdeWUD58kX5/PNruPLKan6HZbJRMCWC14N8L4ezsxdjzsbmzYcYOXIxfftexIoVvSwJ5EIZlghEpDnQAigrIgMDRhXDeQZxmLI2AmOycuDACaZNW8sddzQgNrYM69ffYU8My8UyqxrKDxRxpwn8BhwCrvMyKGOMfz79dD133TWLPXuOceml5aldu7QlgVwuw0Sgqj8AP4jIOFXdHMKYjDE+2LPnKAMGfMfUqWto0KAs06dfY53ERYhgGouPicgLQF0gOvVNVb3Ms6g8YW0ExmQkOTmFli0ns2XLYYYOvZSHHmpCvnxhXANszkgwiWASMBXoinMp6W3AXi+D8pa1ERiTaseOI1xwQWGiovLw6quXERNTjNjYMn6HZUIsmKuGSqvqu0Ciqv6gqv8Cwqw0YIwJlJKivPnmYmrXHsNbby0G4IorqloSiFDBlAgS3f87ReRKYAdQyruQjDFeWrv2D3r3/po5c7Zx+eWV6dKlit8hGZ8FkwiGikhx4H6c+weKAfd5GpUX7C5IY3j33WX07/8t0dFRjBnTiV696lkncSbrRKCqn7svDwLt4M87i8OTfelNBIuJKUaXLlUYMaI9F15YxO9wTA6R2Q1lUcANOH0MfaWqy0WkK/AoUBBoFJoQjTFn6+TJJJ5++hcAhg61TuJM+jIrEbwLVAR+A14TkR1AHPCwqn4SiuCMMWdv7tztxMfPZPXqP/jXv+qhqlYNZNKVWSKIAxqoaoqIRAO7gGqquj80oWU3ayMwkeHIkVMMGvQ/Xn99IRUrFuWrr66lUydrEDYZy+zy0VOqmgKgqieADWeaBESks4isEZH1IvJwBtPcICIrRWSFiLx/JvM/O3ZGZHK3LVsOMWrUEu6+uxHLl99uScBkKbMSQW0RWeq+FqCaOyyAqmqDzGbstjGMADoA24B5IjJdVVcGTFMDeARoqaoJInLeOayLMRErIeEEH364hj59LiI2tgwbNvSmXDlrDDbBySwR1DnHeTcF1qvqBgARmQJ0B1YGTNMbGKGqCQCquuccl2lMxPn443X06/cNe/ceo02bitSqVcqSgDkjmXU6d64dzZUHtgYMbwOapZmmJoCI/ITTtfVgVf0q7YxEpA/QB6BSpUpnF43dR2BymV27jnLPPd8ybdpaGjY8jy+++Ae1atm9nubMBfXweo+XXwNoC1QA5ohIfVU9EDiRqo4GRgPExcWd4xHd2ghM+EtOTqFVq8ls3XqYYcNa8cADcdZJnDlrXiaC7TiXn6aq4L4XaBvwq6omAhtFZC1OYpjnYVzGhK1t2w5TrlwRoqLy8Nprl1GlSnHrKtqcs2A6nUNECopIrTOc9zyghohUEZH8QA9gepppPsEpDSAiZXCqijac4XKMyfVSUpTXX19I7dpjePNNp5O4Ll2qWhIw2SLLRCAiVwGLga/c4YYikvaA/jeqmgT0B2YCq4APVHWFiAwRkW7uZDOB/SKyEvgeeNC7+xSsjcCEp9Wr99O69RQGDPiOSy8tT9euVf0OyeQywVQNDca5Amg2gKouFpGgLkxW1RnAjDTvPRHwWoGB7l9o2J2VJoy8885S+vf/lkKF8jF+fBduuSXW7g422S6obqhV9WCaL5+dXhsTAtWqleCqq6rxxhvtOf/8wn6HY3KpYBLBChG5CYhybwAbAMz1NixjItOJE0kMGfIzAMOGtaJdu0q0a3eWl0wbE6RgGovvwXle8UngfZzuqMPveQRWiDE53E8/badhw/d45plf2bv3GGr3vpgQCaZEUFtVBwGDvA4mNKx+1eQshw+f4tFHf2TEiEVUrlyMmTOvo2PHGL/DMhEkmBLBcBFZJSJPi0g9zyMyJsJs23aYd95Zxj33NGbZsl6WBEzIZZkIVLUdzpPJ9gKjRGSZiDzmeWTG5GL79x//836AOnVKs2HDHbz66mUUKZLf58hMJArqhjJV3aWqrwF9ce4peCKLj+Q8Vt9qcgBVZdq0NcTGjmXAgO9Ys+YPAHtspPFVMDeU1RGRwSKyDOfh9XNxuosIU9ZGYPyxc+cRrr12Otdf/xkVKxZl/vybrZM4kyME01g8BpgKdFLVHR7HY0yu5HQSN4Xt24/w/POt+fe/48ibN6gCuTGeyzIRqGrzUARiTG60deshypcvSlRUHkaMaE+VKsWpWdNKASZnyfCUREQ+cP8vE5GlAX/LAp5cFkasjcCETnJyCq+9dnoncZ06VbEkYHKkzEoE97r/u4YikJCxflqMx1at2k98/Ex+/nkHXbpU4aqrqvkdkjGZyrBEoKo73Zf9VHVz4B/QLzThGRNeRo9eQsOG77F2bQITJlzBF1/8g0qVivkdljGZCqa1qkM673XJ7kCMyQ1q1CjJNddUZ+XKXtx8s/UUasJDhlVDInIXzpl/1TRtAkWBn7wOLPtZG4HJfsePJzJ48FxEhGefbW2dxJmwlFkbwfvAl8AzwMMB7x9W1T88jcpTdoZmssecOVu5446vWbcugb59L0JVrQRgwlJmVUOqqpuAu4HDAX+IiF36YCLWoUMn6ddvFm3aTCU5OYVvv72BN9/sYEnAhK2sSgRdgQU49SqB33IF7Hl5JiLt2HGEceNWMHDgxQwZ0pLCha1/IBPeMkwEqtrV/R/UYylzPOtryJyDffuO8cEHa+jXrxG1a5dm48be9sQwk2sE09dQSxEp7L6+WUReEpHwbQ2z4rs5A6rK1KmriY0dy333fc/atU7zmCUBk5sEc/nom8AxEbkIuB/4HZjgaVTG5AA7dhzh6qs/oZ1Od7wAABnWSURBVEePz6lcuRgLFtxidwabXCmYTueSVFVFpDvwhqq+KyLxXgdmjJ+Sk1No3drpJO7FF9tw770XWydxJtcKJhEcFpFHgFuAViKSB8jnbVhesDYCk7XNmw9SoYLTSdzIkZdTtWpxqlcv6XdYxngqmFOcf+I8uP5fqroL51kEL3galaesjcD8XXJyCi+9NJ86dcb+2Ulcx44xlgRMRAjmUZW7gElAcRHpCpxQ1fc8j8yYEFm+fC8tWrzP/ffPpn37Slx9dQ2/QzImpIK5augG4DfgeuAG4FcRuc7rwIwJhbfeWkzjxhPYsOEg779/JdOnX0OFCkX9DsuYkAqmjWAQ0ERV9wCISFngG2Cal4FlP2sjMH9J7Q6iTp3SXH99LV55pR1lyxbyOyxjfBFMIsiTmgRc+wnyofc5k7URRLJjxxJ54omfiIoSnnuuDW3aVKRNm4p+h2WMr4I5oH8lIjNFpJeI9AK+AGZ4G5Yx2W/27C00aDCe4cPnc+RIImp3mxsDBPfM4gdF5B/Ape5bo1X1Y2/DMib7HDx4koce+oHRo5dSrVoJvvvuBusq2pgAmT2PoAbwIlANWAY8oKrbQxVYtrOzv4i1c+cRJk5cyQMPxPHUUy0pVCgMb4MxxkOZVQ2NAT4HrsXpgfT1M525iHQWkTUisl5EHs5kumtFREUk7kyXccasr6GIsHfvMV5/fSEAtWuXZtOmPrzwQltLAsakI7OqoaKq+rb7eo2ILDyTGYtIFDAC51GX24B5IjJdVVemma4ocC/w65nM35j0qCqTJ69mwIDvOHToJJ06xVCzZim7IsiYTGRWIogWkUYi0lhEGgMF0wxnpSmwXlU3qOopYArQPZ3pngaeA06ccfTGBNi69RBXXfUxPXt+QfXqJVi06FbrJM6YIGRWItgJvBQwvCtgWIHLsph3eWBrwPA2oFngBG5CqaiqX4jIgxnNSET6AH0AKlU620Y+ayPIzZKSUmjbdiq7dh3l5Zfbcc89jYiKCuOrnI0JocweTNPOywW7nde9BPTKalpVHQ2MBoiLizvHI7q1EeQmmzYdpGLFouTNm4dRozpStWpxqlYt4XdYxoQVL0+ZtgOBd+pUcN9LVRSoB8wWkU3AJcD0kDQYm7CXlJTCiy/Oo06dsYwc6XQSd/nllS0JGHMWgrmz+GzNA2qISBWcBNADuCl1pKoeBMqkDovIbJxLVOd7GJPJBZYu3Ut8/FfMn7+b7t2rc+21Nf0OyZiw5lkiUNUkEekPzASigDGqukJEhgDzVXW6V8vOIKLQLs54YuTIRdx77/eULFmAqVO7cv31tRC7JNiYc5JlIhDnV9YTqKqqQ9znFV+gqr9l9VlVnUGa7ihU9YkMpm0bVMTnzA4a4Si1k7h69crQo0dtXn65LWXK2CWhxmSHYEoEI4EUnKuEhgCHgY+AJh7GZQwAR4+e4rHHfiJvXuGFF9rSunVFWre2TuKMyU7BNBY3U9W7ca/zV9UEIL+nURkDfPvtZurXH88rryzg5Mlk6yTOGI8EUyJIdO8SVvjzeQQpnkblBTuIhI0DB07wwAM/8O67y6hRoyRz5vSgVasKfodlTK4VTIngNeBj4DwR+Q/wP2CYp1F5yRoWc7zdu48xZcpq/u//mrJkya2WBIzxWDDdUE8SkQVAe5yW1qtVdZXnkZmIsnv3UaZMWc29915MrVql2LSptzUGGxMiwVw1VAk4BnwW+J6qbvEysOxnVUM5kaoyadIq7r33O44cSeSKK6pSo0ZJSwLGhFAwbQRf4BxFBYgGqgBrgLoexmUiwJYth+jbdxZffrmR5s3L8e67nahRo6TfYRkTcYKpGqofOOx2FNfPs4g8Z20EOUFqJ3F79hzjtdcuo1+/htZJnDE+OeM7i1V1oYg0y3pKY/5uw4YDVK5cjLx58/D22x2pVq0EMTHF/Q7LmIgWTBvBwIDBPEBjYIdnEXnG2gj8lJSUwvDh83jyybk8/3wbBgxoTPv2lf0OyxhDcCWCogGvk3DaDD7yJpxQsKqhUFu8eA/x8TNZuHA311xTg+uvt07ijMlJMk0E7o1kRVX1gRDFY3KZN95YyL//PZvSpaOZNq2b9RRqTA6UYSIQkbxuD6ItQxmQyR1SO4lr0KAsPXvW4aWX2lKqVEG/wzLGpCOzEsFvOO0Bi0VkOvAhcDR1pKr+1+PYspd1MRESR46cYtCg/5EvXx5efNE6iTMmHARzvV40sB+n99GuwFXu//BkXUx45uuvN1Gv3jhef30hiYkp1kmcMWEisxLBee4VQ8v564ayVPYLN39KSDjBwIHfM27cCmrVKsWcOT249FLrH8iYcJFZIogCipD+ZTaWCMyf9uw5xrRpa3nkkWY88URzoqO9fAKqMSa7ZfaL3amqQ0IWiecsd2WnXbuOMnnyKv797zi3k7g+lC5tjcHGhKPM2ghyaWV6Ll2tEFFVxo9fTmzsWB555EfWrUsAsCRgTBjLLBG0D1kUJixs2nSQzp0/olevr4iNLc3ixbdaJ3HG5AIZVg2p6h+hDMTkbElJKbRrN5V9+44zYkR7+vZtSJ48VroyJjeIoFY9ayM4G+vXJ1ClSnHy5s3DmDGdqVq1OJUrWydxxuQmEdjvr53FBiMxMZlhw36hbt1xjBixGIB27SpZEjAmF4qgEoEJ1sKFu4mPn8nixXu4/vqa/POftfwOyRjjIUsE5jSvvbaQgQO/p2zZQvz3v9255poafodkjPFY5CQC6+4gU6mdxDVqdB633lqX4cPbUrJktN9hGWNCIHISQSrra+g0hw+f4pFH5lCgQBTDh7ejVasKtGpl3UMYE0kisLHYpPrqq43UqzeWkSMXo4p1EmdMhIq8EoFh//7jDBz4Pe+9t5I6dUrx00830bx5Ob/DMsb4JIISgZ3tptq//zgff7yexx+/hEGDLqFAgQj6Ghhj/sbTqiER6Swia0RkvYg8nM74gSKyUkSWisi3IhKCp5lHZhvBzp1HePHFeagqNWuWYvPmPgwZcqklAWOMd4nAfd7xCKALEAvcKCKxaSZbBMSpagNgGvC8V/FEKlVlzJhl1Kkzlscf/4n16w8A2BVBxpg/eVkiaAqsV9UNqnoKmAJ0D5xAVb9X1WPu4C+AXa6SjTZuPEDHjtOIj5/JRReVZckS6yTOGPN3XtYLlAe2BgxvA5plMn088GV6I0SkD9AHoFKlSmcXTYRdEZOUlMJll33A/v0nePPNy+nT5yLrJM4Yk64cUUEsIjcDcUCb9Mar6mhgNEBcXNw5HtFz98Fw3boEqlZ1OokbO7Yz1aqVoGLFYn6HZYzJwbysGtoOVAwYruC+dxoRuRwYBHRT1ZMexpOrJSYmM3Toz9SrN4433lgEQNu2lSwJGGOy5GWJYB5QQ0Sq4CSAHsBNgROISCNgFNBZVfd4GEuuNn/+LuLjZ7J06V569KjNjTfW9jskY0wY8SwRqGqSiPQHZgJRwBhVXSEiQ4D5qjodeAEoAnwoTtcPW1S1m0cReTNbn7366gIGDpzNBRcU5tNPr6Zbt+p+h2SMCTOethGo6gxgRpr3ngh4fbmXy09XLulrKLWTuLi4C4iPr8/zz7emRAm7JNQYc+ZyRGOxCd6hQyf5v/+bQ3R0Xl5+uR0tW5anZcvyfodljAlj1ulcGJkxYwN1645j9Oil5M0r1kmcMSZbRFCJIHwPmvv2HeO++75n0qRV1K1bmmnTbqJZswv9DssYk0tEUCJIFX5tBAkJJ/nss9958snmPProJeTPH+V3SMaYXCQCE0F42L79MJMmreLBB5tQo0ZJNm/uY43BxhhPWBtBDqOqvP32UmJjxzJ48Fx+/93pJM6SgDHGK5GTCMKgYfX33w/Qvv0H9OnzNY0bn8/SpbdRvbp1EmeM8VbkVQ3l0PsIkpJSaN/+A/744wSjRnXgjjsaWCdxxpiQiLxEkMOsWfMH1aqVIG/ePIwf34Vq1UpQoUJRv8MyxkSQyKkaymFOnUrmqafmUr/+OEaMcDqJa9OmoiUBY0zIRVCJIOe0Efz2207i42eyfPk+brqpDj171vE7JGNMBIugRJDK33r3V15ZwP33z+bCCwvz2WfX0LVrNV/jMcaYCEwE/kjtJK5p0wvo3bsBzz3XmuLFC/gdljHGWCLw2sGDJ3nooR8oWDAvr7xyGS1alKdFC+skzhiTc0RQY3Ho2wg+++x3YmPH8s47yyhQIMo6iTPG5EgRWCLwvo1g795j3Hvvd0yevJr69cvwySfdadLEOokzxuRMEZgIvHfw4ElmzNjIU0+14OGHm1knccaYHM0SQTbZuvUQEyeu4uGHm1K9utNJnDUGG2PCQeS0EXhUP5+Sorz11mLq1h3H0KE//9lJnCUBY0y4iJxEkCob+xpaty6Byy6byl13fUPTphewbFkv6yTOGBN2rGroLCUlpdChw4ccOHCSd9/txO2310NyaId2xhiTGUsEZ2jVqv3UqFGSvHnzMGHCFVSrVoJy5Yr4HZYxJgiJiYls27aNEydO+B2KZ6Kjo6lQoQL58uUL+jMRlAjOrY3g5Mkkhg37lWHDfuWFF9pw330X06pVhWyKzRgTCtu2baNo0aLExMTkyhK8qrJ//362bdtGlSpVgv5cBCWCVGe+83/5ZQfx8TNZuXI/t9wSyy23xHoQlzHGaydOnMi1SQBARChdujR79+49o89FYCI4M8OHz+PBB3+gQoWizJjxD7p0qep3SMaYc5Bbk0Cqs1k/SwQZSElR8uQRmjcvR9++F/Hss60pVswuCTXG5D4RdPlocG0EBw6cID7+K+699zsAWrQoz8iRHSwJGGOyRVRUFA0bNqRevXpcddVVHDhw4M9xK1as4LLLLqNWrVrUqFGDp59++rQ+yr788kvi4uKIjY2lUaNG3H///dkSUwQlglQZF5s++WQdsbFjGT9+BUWL5rdO4owx2a5gwYIsXryY5cuXU6pUKUaMGAHA8ePH6datGw8//DBr1qxhyZIlzJ07l5EjRwKwfPly+vfvz8SJE1m5ciXz58+nevXq2RKTVQ0Be/YcpX//b/nww7U0bHgen3/+Dxo3Pt/vsIwxXlpwHyQszt55lmwIF78S9OTNmzdn6dKlALz//vu0bNmSjh07AlCoUCHeeOMN2rZty913383zzz/PoEGDqF27NuCULO66665sCTsCSwR/d+jQKWbN2sx//nMpv/3W05KAMcZzycnJfPvtt3Tr1g1wqoUuvvji06apVq0aR44c4dChQyxfvvxv47NL5JQI0lTzbNlyiAkTVvLoo82oXr0kW7bcSdGi+X0KzhgTcmdw5p6djh8/TsOGDdm+fTt16tShQ4cOvsQRyNMSgYh0FpE1IrJeRB5OZ3wBEZnqjv9VRGK8jAcgRWHkyEXUrTuWYcN++bOTOEsCxphQSG0j2Lx5M6r6ZxtBbGwsCxYsOG3aDRs2UKRIEYoVK0bdunX/Nj67eJYIRCQKGAF0AWKBG0Uk7Z1Y8UCCqlYHXgae8yoegDU7ytK2/afcffe3NG9ejhUrbrdO4owxvihUqBCvvfYaw4cPJykpiZ49e/K///2Pb775BnBKDgMGDOChhx4C4MEHH2TYsGGsXbsWgJSUFN56661sicXLEkFTYL2qblDVU8AUoHuaaboD493X04D24tHdHklJ0Om5O1i2fD9jx3Zm5szriIkp7sWijDEmKI0aNaJBgwZMnjyZggUL8umnnzJ06FBq1apF/fr1adKkCf379wegQYMGvPLKK9x4443UqVOHevXqsWHDhmyJw8s2gvLA1oDhbUCzjKZR1SQROQiUBvYFTiQifYA+AJUqVTqrYPKWqsXEpz6hWrf7ubBC6bOahzHGnKsjR46cNvzZZ5/9+bp+/frMnj07w8927dqVrl27ZntMYdFYrKqjgdEAcXFxZ3dxf4VuXNqvW3aGZYwxuYKXVUPbgYoBwxXc99KdRkTyAsWB/R7GZIwxJg0vE8E8oIaIVBGR/EAPYHqaaaYDt7mvrwO+U7ud1xjjodx+iDmb9fMsEahqEtAfmAmsAj5Q1RUiMkREUuto3gVKi8h6YCDwt0tMjTEmu0RHR7N///5cmwxSn0cQHR19Rp+TcNsgcXFxOn/+fL/DMMaEoUh+QpmILFDVuPQ+ExaNxcYYkx3y5ct3Rk/uihTW15AxxkQ4SwTGGBPhLBEYY0yEC7vGYhHZC2w+y4+XIc1dyxHA1jky2DpHhnNZ58qqWja9EWGXCM6FiMzPqNU8t7J1jgy2zpHBq3W2qiFjjIlwlgiMMSbCRVoiGO13AD6wdY4Mts6RwZN1jqg2AmOMMX8XaSUCY4wxaVgiMMaYCJcrE4GIdBaRNSKyXkT+1qOpiBQQkanu+F9FJCb0UWavINZ5oIisFJGlIvKtiFT2I87slNU6B0x3rYioiIT9pYbBrLOI3ODu6xUi8n6oY8xuQXy3K4nI9yKyyP1+X+FHnNlFRMaIyB4RWZ7BeBGR19ztsVREGp/zQlU1V/0BUcDvQFUgP7AEiE0zTT/gLfd1D2Cq33GHYJ3bAYXc13dFwjq70xUF5gC/AHF+xx2C/VwDWASUdIfP8zvuEKzzaOAu93UssMnvuM9xnVsDjYHlGYy/AvgSEOAS4NdzXWZuLBE0Bdar6gZVPQVMAbqnmaY7MN59PQ1oLyISwhizW5brrKrfq+oxd/AXnCfGhbNg9jPA08BzQG7odziYde4NjFDVBABV3RPiGLNbMOusQDH3dXFgRwjjy3aqOgf4I5NJugPvqeMXoISIXHguy8yNiaA8sDVgeJv7XrrTqPMAnYNAOD/RPph1DhSPc0YRzrJcZ7fIXFFVvwhlYB4KZj/XBGqKyE8i8ouIdA5ZdN4IZp0HAzeLyDZgBnBPaELzzZn+3rNkzyOIMCJyMxAHtPE7Fi+JSB7gJaCXz6GEWl6c6qG2OKW+OSJSX1UP+BqVt24ExqnqcBFpDkwQkXqqmuJ3YOEiN5YItgMVA4YruO+lO42I5MUpTu4PSXTeCGadEZHLgUFAN1U9GaLYvJLVOhcF6gGzRWQTTl3q9DBvMA5mP28DpqtqoqpuBNbiJIZwFcw6xwMfAKjqz0A0TudsuVVQv/czkRsTwTyghohUEZH8OI3B09NMMx24zX19HfCduq0wYSrLdRaRRsAonCQQ7vXGkMU6q+pBVS2jqjGqGoPTLtJNVcP5OafBfLc/wSkNICJlcKqKNoQyyGwWzDpvAdoDiEgdnESwN6RRhtZ04Fb36qFLgIOquvNcZpjrqoZUNUlE+gMzca44GKOqK0RkCDBfVacD7+IUH9fjNMr08C/icxfkOr8AFAE+dNvFt6hqN9+CPkdBrnOuEuQ6zwQ6ishKIBl4UFXDtrQb5DrfD7wtIv/GaTjuFc4ndiIyGSeZl3HbPZ4E8gGo6ls47SBXAOuBY8Dt57zMMN5exhhjskFurBoyxhhzBiwRGGNMhLNEYIwxEc4SgTHGRDhLBMYYE+EsEZgcSUSSRWRxwF9MJtMeyYbljRORje6yFrp3qJ7pPN4RkVj39aNpxs091xjd+aRul+Ui8pmIlMhi+obh3hun8Z5dPmpyJBE5oqpFsnvaTOYxDvhcVaeJSEfgRVVtcA7zO+eYspqviIwH1qrqfzKZvhdOr6v9szsWk3tYicCEBREp4j5HYaGILBORv/U0KiIXisicgDPmVu77HUXkZ/ezH4pIVgfoOUB197MD3XktF5H73PcKi8gXIrLEff+f7vuzRSRORJ4FCrpxTHLHHXH/TxGRKwNiHici14lIlIi8ICLz3D7m7wxis/yM29mYiDR113GRiMwVkVrunbhDgH+6sfzTjX2MiPzmTptej60m0vjd97b92V96fzh3xS52/z7GuQu+mDuuDM5dlakl2iPu//uBQe7rKJz+hsrgHNgLu+//H/BEOssbB1znvr4e+BW4GFgGFMa5K3sF0Ai4Fng74LPF3f+zcZ95kBpTwDSpMV4DjHdf58fpRbIg0Ad4zH2/ADAfqJJOnEcC1u9DoLM7XAzI676+HPjIfd0LeCPg88OAm93XJXD6Iirs9/62P3//cl0XEybXOK6qDVMHRCQfMExEWgMpOGfC5wO7Aj4zDxjjTvuJqi4WkTY4Dyv5ye1aIz/OmXR6XhCRx3D6qYnH6b/mY1U96sbwX6AV8BUwXESew6lO+vEM1utL4FURKQB0Buao6nG3OqqBiFznTlccp7O4jWk+X1BEFrvrvwqYFTD9eBGpgdPNQr4Mlt8R6CYiD7jD0UAld14mQlkiMOGiJ1AWuFhVE8XpUTQ6cAJVneMmiiuBcSLyEpAAzFLVG4NYxoOqOi11QETapzeRqq4V51kHVwBDReRbVR0SzEqo6gkRmQ10Av6J86AVcJ42dY+qzsxiFsdVtaGIFMLpf+du4DWcB/B8r6rXuA3rszP4vADXquqaYOI1kcHaCEy4KA7scZNAO+Bvz1wW5znMu1X1beAdnMf9/QK0FJHUOv/CIlIzyGX+CFwtIoVEpDBOtc6PIlIOOKaqE3E680vvmbGJbskkPVNxOgpLLV2Ac1C/K/UzIlLTXWa61Hna3ADgfvmrK/XUroh7BUx6GKeKLNVM4B5xi0fi9EprIpwlAhMuJgFxIrIMuBVYnc40bYElIrII52z7VVXdi3NgnCwiS3GqhWoHs0BVXYjTdvAbTpvBO6q6CKgP/OZW0TwJDE3n46OBpamNxWl8jfNgoG/UefwiOIlrJbBQnIeWjyKLErsby1KcB7M8Dzzjrnvg574HYlMbi3FKDvnc2Fa4wybC2eWjxhgT4axEYIwxEc4SgTHGRDhLBMYYE+EsERhjTISzRGCMMRHOEoExxkQ4SwTGGBPh/h/j42dqb2jDOAAAAABJRU5ErkJggg==\n"},"metadata":{"needs_background":"light"}},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["###AdaBoost"],"metadata":{"id":"5HfvEig4LWH5"}},{"cell_type":"code","source":["from sklearn.ensemble import AdaBoostClassifier\n","model_adasmt = AdaBoostClassifier(DecisionTreeClassifier(max_depth=2),\n","                                  n_estimators=50,\n","                                  algorithm=\"SAMME.R\",\n","                                  learning_rate=0.5)\n","model_adasmt, accuracy_adasmt, roc_auc_adasmt, f1_score_adasmt,coh_kap_adasmt, tt_adasmt = run_model(model_adasmt, X_train_smt, y_train_smt, X_test, y_test)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":800},"id":"728lsl4hLlAq","outputId":"7a3939da-c513-4ea1-c6ba-b9a69494e324","executionInfo":{"status":"ok","timestamp":1659224789722,"user_tz":-420,"elapsed":248165,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":32,"outputs":[{"output_type":"stream","name":"stdout","text":["Accuracy = 0.9898528843790597\n","ROC Area under Curve = 0.9207919050817165\n","F1 Score  = 0.22933333333333333\n","Cohen's Kappa = 0.22696113551593144\n","Time taken = 247.0561490058899\n","              precision    recall  f1-score   support\n","\n","           0    0.99973   0.99010   0.99489     56861\n","           1    0.13251   0.85149   0.22933       101\n","\n","    accuracy                        0.98985     56962\n","   macro avg    0.56612   0.92079   0.61211     56962\n","weighted avg    0.99820   0.98985   0.99354     56962\n","\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["### TabNet"],"metadata":{"id":"Z77N4FLAR2h_"}},{"cell_type":"code","source":["!pip install pytorch-tabnet"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"yb2_fSQrSC68","executionInfo":{"status":"ok","timestamp":1659224792225,"user_tz":-420,"elapsed":2511,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}},"outputId":"7a32bf35-3432-499c-e726-902dcb4c0d56"},"execution_count":33,"outputs":[{"output_type":"stream","name":"stdout","text":["Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n","Requirement already satisfied: pytorch-tabnet in /usr/local/lib/python3.7/dist-packages (3.1.1)\n","Requirement already satisfied: scipy>1.4 in /usr/local/lib/python3.7/dist-packages (from pytorch-tabnet) (1.7.3)\n","Requirement already satisfied: tqdm<5.0,>=4.36 in /usr/local/lib/python3.7/dist-packages (from pytorch-tabnet) (4.64.0)\n","Requirement already satisfied: scikit_learn>0.21 in /usr/local/lib/python3.7/dist-packages (from pytorch-tabnet) (1.0.2)\n","Requirement already satisfied: numpy<2.0,>=1.17 in /usr/local/lib/python3.7/dist-packages (from pytorch-tabnet) (1.21.6)\n","Requirement already satisfied: torch<2.0,>=1.2 in /usr/local/lib/python3.7/dist-packages (from pytorch-tabnet) (1.12.0+cu113)\n","Requirement already satisfied: joblib>=0.11 in /usr/local/lib/python3.7/dist-packages (from scikit_learn>0.21->pytorch-tabnet) (1.1.0)\n","Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.7/dist-packages (from scikit_learn>0.21->pytorch-tabnet) (3.1.0)\n","Requirement already satisfied: typing-extensions in /usr/local/lib/python3.7/dist-packages (from torch<2.0,>=1.2->pytorch-tabnet) (4.1.1)\n"]}]},{"cell_type":"code","source":["import torch\n","import torch.nn as nn\n","import torch.optim as optim\n","from torch.optim.lr_scheduler import ReduceLROnPlateau\n","import torch.nn.functional as F\n","\n","import random\n","\n","from pytorch_tabnet.pretraining import TabNetPretrainer\n","from pytorch_tabnet.tab_model import TabNetClassifier\n","from pytorch_tabnet.metrics import Metric\n","from pathlib import Path"],"metadata":{"id":"zWRTChTwRhVe","executionInfo":{"status":"ok","timestamp":1659224792795,"user_tz":-420,"elapsed":585,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":34,"outputs":[]},{"cell_type":"code","source":["class F1Score(Metric):\n","    def __init__(self):\n","        self._name = \"f1_score\"\n","        self._maximize = True\n","\n","    def __call__(self, y_true, y_pred):\n","        return f1_score(y_true, np.argmax(y_pred, axis = 1))"],"metadata":{"id":"iciMt6SYSBKU","executionInfo":{"status":"ok","timestamp":1659224792796,"user_tz":-420,"elapsed":5,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":35,"outputs":[]},{"cell_type":"code","source":["tabnet_params = dict(n_d=64, n_a=64, n_steps=5,\n","    gamma=1.5, n_independent=2, n_shared=2,\n","    cat_idxs=[],\n","    cat_dims = [],\n","    cat_emb_dim=1,\n","    lambda_sparse=1e-4, momentum=0.3, clip_value=2.,\n","    optimizer_fn=torch.optim.Adam,\n","    optimizer_params=dict(lr=2e-2, weight_decay = 1e-5),\n","    scheduler_params = dict(\n","        mode = \"max\", patience = 5, min_lr = 1e-5, factor = 0.9),                 \n","    scheduler_fn=torch.optim.lr_scheduler.ReduceLROnPlateau, \n","    mask_type='entmax',\n","    verbose =10\n",")"],"metadata":{"id":"chcvEMzvTqVd","executionInfo":{"status":"ok","timestamp":1659224792798,"user_tz":-420,"elapsed":6,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":36,"outputs":[]},{"cell_type":"code","source":["import itertools\n","\n","# Create a confusion matrix\n","def custom_plot_confusion_matrix(cm, classes,\n","                          normalize=False,\n","                          title='Confusion matrix',\n","                          cmap=plt.cm.Blues):\n","    \"\"\"\n","    This function prints and plots the confusion matrix.\n","    Normalization can be applied by setting `normalize=True`.\n","    \"\"\"\n","    if normalize:\n","        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n","        print(\"Normalized confusion matrix\")\n","    else:\n","        print('Confusion matrix, without normalization')\n","\n","    print(cm)\n","\n","    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n","    plt.title(title, fontsize=14)\n","    plt.colorbar()\n","    tick_marks = np.arange(len(classes))\n","    plt.xticks(tick_marks, classes, rotation=45)\n","    plt.yticks(tick_marks, classes)\n","\n","    fmt = '.2f' if normalize else 'd'\n","    thresh = cm.max() / 2.\n","    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n","        plt.text(j, i, format(cm[i, j], fmt),\n","                 horizontalalignment=\"center\",\n","                 color=\"white\" if cm[i, j] > thresh else \"black\")\n","\n","    plt.tight_layout()\n","    plt.ylabel('True label')\n","    plt.xlabel('Predicted label')"],"metadata":{"id":"6lIm-nTElgvz","executionInfo":{"status":"ok","timestamp":1659226432805,"user_tz":-420,"elapsed":474,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":42,"outputs":[]},{"cell_type":"code","source":["def run_tabnet_model(model, X_train, y_train, X_test, y_test):\n","\n","    t0=time.time()\n","\n","    model.fit(X_train=X_train, y_train=y_train,\n","        eval_set=[(X_train, y_train), (X_test, y_test)],\n","        eval_name=['train', 'valid'],\n","        max_epochs = 200,\n","        patience = 50,\n","        batch_size = 1024 * 15, \n","        virtual_batch_size = 256 * 10,\n","        num_workers = 4,\n","        drop_last = False,\n","        eval_metric=['f1_score']\n","        )\n","    \n","    y_pred = model.predict(X_test)\n","    accuracy = accuracy_score(y_test, y_pred)\n","    roc_auc = roc_auc_score(y_test, y_pred)\n","    precision = precision_score(y_test, y_pred)\n","    recall = recall_score(y_test, y_pred)\n","    f1 = f1_score(y_test, y_pred)\n","    coh_kap = cohen_kappa_score(y_test, y_pred)\n","    time_taken = time.time()-t0\n","    print(\"Accuracy = {}\".format(accuracy))\n","    print(\"ROC Area under Curve = {}\".format(roc_auc))\n","    #print(\"Precision  = {}\".format(precision))\n","    #print(\"Recall  = {}\".format(recall))\n","    print(\"F1 Score  = {}\".format(f1))\n","    print(\"Cohen's Kappa = {}\".format(coh_kap))\n","    print(\"Time taken = {}\".format(time_taken))\n","    print(classification_report(y_test,y_pred,digits=5))\n","    \n","    probs = model.predict_proba(X_test)  \n","    probs = probs[:, 1]  \n","    fper, tper, thresholds = roc_curve(y_test, probs) \n","    plot_roc_cur(fper, tper)\n","    \n","    oversample_cm = confusion_matrix(y_test, y_pred)\n","    custom_plot_confusion_matrix(oversample_cm, classes = [0, 1], title=\"TabNet + SMOTE \\n Confusion Matrix\", cmap=plt.cm.Blues)\n","    \n","    return model, accuracy, roc_auc, f1, coh_kap, time_taken"],"metadata":{"id":"PDJKstW0UrVO","executionInfo":{"status":"ok","timestamp":1659226548371,"user_tz":-420,"elapsed":310,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":47,"outputs":[]},{"cell_type":"code","source":["model_tabnetsmt =  TabNetClassifier(**tabnet_params)\n","\n","model_tabnetsmt, accuracy_tabnetsmt, roc_auc_tabnetsmt, f1_score_tabnetsmt,coh_kap_tabnetsmt, tt_tabnetsmt = run_tabnet_model(model_tabnetsmt, X_train_smt, y_train_smt, X_test, y_test)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000},"id":"lvExxME8T2jT","executionInfo":{"status":"ok","timestamp":1659227672324,"user_tz":-420,"elapsed":1121554,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}},"outputId":"a1af6ba1-4ded-4b6d-9d89-41e99dafeba6"},"execution_count":48,"outputs":[{"output_type":"stream","name":"stdout","text":["Device used : cuda\n","epoch 0  | loss: 0.28386 | train_f1_score: 0.74698 | valid_f1_score: 0.00522 |  0:00:05s\n","epoch 10 | loss: 0.00709 | train_f1_score: 0.99853 | valid_f1_score: 0.47911 |  0:01:18s\n","epoch 20 | loss: 0.00412 | train_f1_score: 0.99936 | valid_f1_score: 0.60215 |  0:02:17s\n","epoch 30 | loss: 0.00666 | train_f1_score: 0.99814 | valid_f1_score: 0.4607  |  0:03:16s\n","epoch 40 | loss: 0.00229 | train_f1_score: 0.99952 | valid_f1_score: 0.65587 |  0:04:15s\n","epoch 50 | loss: 0.00139 | train_f1_score: 0.99979 | valid_f1_score: 0.67556 |  0:05:15s\n","epoch 60 | loss: 0.00288 | train_f1_score: 0.99965 | valid_f1_score: 0.68313 |  0:06:15s\n","epoch 70 | loss: 0.00097 | train_f1_score: 0.9997  | valid_f1_score: 0.72115 |  0:07:15s\n","epoch 80 | loss: 0.00107 | train_f1_score: 0.99987 | valid_f1_score: 0.72321 |  0:08:15s\n","epoch 90 | loss: 0.00221 | train_f1_score: 0.99945 | valid_f1_score: 0.6171  |  0:09:15s\n","epoch 100| loss: 0.00148 | train_f1_score: 0.99975 | valid_f1_score: 0.72961 |  0:10:10s\n","epoch 110| loss: 0.00087 | train_f1_score: 0.99991 | valid_f1_score: 0.76923 |  0:11:05s\n","epoch 120| loss: 0.0007  | train_f1_score: 0.99986 | valid_f1_score: 0.71681 |  0:12:00s\n","epoch 130| loss: 0.00076 | train_f1_score: 0.99988 | valid_f1_score: 0.72247 |  0:12:55s\n","epoch 140| loss: 0.00053 | train_f1_score: 0.99993 | valid_f1_score: 0.83    |  0:13:50s\n","epoch 150| loss: 0.00085 | train_f1_score: 0.99986 | valid_f1_score: 0.75    |  0:14:46s\n","epoch 160| loss: 0.00038 | train_f1_score: 0.99996 | valid_f1_score: 0.78469 |  0:15:42s\n","epoch 170| loss: 0.00044 | train_f1_score: 0.99995 | valid_f1_score: 0.79612 |  0:16:39s\n","epoch 180| loss: 0.00203 | train_f1_score: 0.99963 | valid_f1_score: 0.70293 |  0:17:36s\n","epoch 190| loss: 0.00049 | train_f1_score: 0.99996 | valid_f1_score: 0.80788 |  0:18:37s\n","\n","Early stopping occurred at epoch 190 with best_epoch = 140 and best_valid_f1_score = 0.83\n","Best weights from best epoch are automatically used!\n","Accuracy = 0.999403110845827\n","ROC Area under Curve = 0.9107503951358888\n","F1 Score  = 0.83\n","Cohen's Kappa = 0.8297010605147443\n","Time taken = 1120.540057182312\n","              precision    recall  f1-score   support\n","\n","           0    0.99968   0.99972   0.99970     56861\n","           1    0.83838   0.82178   0.83000       101\n","\n","    accuracy                        0.99940     56962\n","   macro avg    0.91903   0.91075   0.91485     56962\n","weighted avg    0.99940   0.99940   0.99940     56962\n","\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}},{"output_type":"stream","name":"stdout","text":["Confusion matrix, without normalization\n","[[56845    16]\n"," [   18    83]]\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["##Model Selection"],"metadata":{"id":"kkD3spsJ-jd0"}},{"cell_type":"code","source":["accuracy_scores = [accuracy_lrsmt, accuracy_dtsmt, accuracy_rfsmt, accuracy_lgbsmt, accuracy_cbsmt, accuracy_xgbsmt, accuracy_adasmt, accuracy_tabnetsmt]\n","roc_auc_scores = [roc_auc_lrsmt, roc_auc_dtsmt, roc_auc_rfsmt, roc_auc_lgbsmt, roc_auc_cbsmt, roc_auc_xgbsmt, roc_auc_adasmt, roc_auc_tabnetsmt]\n","f1_scores = [f1_score_lrsmt, f1_score_dtsmt, f1_score_rfsmt, f1_score_lgbsmt, f1_score_cbsmt, f1_score_xgbsmt, f1_score_adasmt, f1_score_tabnetsmt]\n","coh_kap_scores = [coh_kap_lrsmt, coh_kap_dtsmt,coh_kap_rfsmt, coh_kap_lgbsmt, coh_kap_cbsmt, coh_kap_xgbsmt, coh_kap_adasmt, coh_kap_tabnetsmt]\n","tt = [tt_lrsmt, tt_dtsmt, tt_rfsmt, tt_lgbsmt, tt_cbsmt, tt_xgbsmt, tt_adasmt, tt_tabnetsmt]\n","\n","model_data = {'Model': ['Logistic Regression','Decision Tree','Random Forest','LightGBM','Catboost','XGBoost', 'AdaBoost', 'TabNet'],\n","              'Accuracy': accuracy_scores,\n","              'ROC_AUC': roc_auc_scores,\n","              'F1 Score': f1_scores,\n","              'Cohen_Kappa': coh_kap_scores,\n","              'Time taken': tt}\n","data = pd.DataFrame(model_data)\n","\n","fig, ax1 = plt.subplots(figsize=(12,10))\n","ax1.set_title('Model Comparison: Accuracy and Time taken for execution', fontsize=13)\n","color = 'tab:green'\n","ax1.set_xlabel('Model', fontsize=13)\n","ax1.set_ylabel('Time taken', fontsize=13, color=color)\n","ax2 = sns.barplot(x='Model', y='Time taken', data = data, palette='summer')\n","ax1.tick_params(axis='y')\n","ax2 = ax1.twinx()\n","color = 'tab:red'\n","ax2.set_ylabel('F1 Score', fontsize=13, color=color)\n","ax2 = sns.lineplot(x='Model', y='F1 Score', data = data, sort=False, color=color)\n","ax2.tick_params(axis='y', color=color)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":625},"id":"N5YlxUQ0NxqM","outputId":"3c8408e9-4689-47b1-a910-27e7ad8255da","executionInfo":{"status":"ok","timestamp":1659227698651,"user_tz":-420,"elapsed":12,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":50,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 864x720 with 2 Axes>"],"image/png":"iVBORw0KGgoAAAANSUhEUgAAAwUAAAJgCAYAAADbK88kAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdeXxU5d3//9dnJslkZU9ANgk7VFsX6kZ6/6wrWottsb2ltWqlLtUWi1rFu7ba+m3FulC1WqXKLTcutRarFvet3nW73W69rWRGhh0EAoEEkjCT7fr9MSc4RAIJJDmZmffz8ZhHMuecOec9Z84k85lzXdcx5xwiIiIiIpK5An4HEBERERERf6koEBERERHJcCoKREREREQynIoCEREREZEMp6JARERERCTDqSgQEREREclwKgpEfGRmjWZ2bDuXPdbMGrs4Uo9hZjVmdrTfOaTzmdm9Znb/Pj72K2ZW1cmROp2ZXWdmL3bj9kaa2Rtmts3M/tZd2+1pzOwZM7vS7xwiqUhFgcgemNk/zMyZ2XdaTT/Sm77Sp2i7ZWa9zOx3ZrbUzGrNbJ2ZPWVmx/udraOcc4XOuTf9zpHMzH7uve7n+J0lHZnZcK8YbLk1mll90v2PnXP/dM716eZc3foBfx/NBtYAvZ1z3/Q7THfw3otlydOcc6c4537nVyaRVKaiQGTvyoHzW00735veY5hZIfAa8BXgu0BfYBQwDzjDx2gdYmbZfmfYHTMLkHjdtwAX+JijR+6fzuCcW+0Vg4XOuULgH8Bvk6Z9weeIPdlI4CO3j1ckTefjSkTaR0WByN49BhxqZiMBzKwImAb8Z/JCZpZvZreZ2Roz22xmj5vZ8KT5RWa2wMy2mNmq3X3bbGbfMLP3zKzKzMrN7HsdyPlTYDBwqnPuHedcvXMu5px7wjn3o6RtTDOzD82s2vv5zaR555pZ1MxmmdlaM9tuZjebWX8zW+Q1TQgnfztnZveb2YNmttCbv8zMzk2aP9TMnjWzTd42/2lmhyfNv87MXva2sxF40pu+81tAMxthZs95+2Wrmb1vZuO8eVlm9kszW+7Ne8nMDmqVb6GZ/cl7/Dozu7AD+7XFycAQ4GzgmORteNv5YtLz3JL8zbKX/1EzW+9leN3M+rd+nt79XZqJeWerfu8dT9uAy/e2T73HfcvM3vW2t8HMfmNmQe/5f7PVsv9lZvft7kmb2ZfM7FXvmN5qieYZozqyf83sPO+42GZmC4Hcdu/1z+dpvX9atj8/afvTzewQM3vHO4ZfMbPBSY/J9463Fd5r9ayZjW5je/8O/AdwrH12xmJke16DVuv5gfeeOtK73+Z73T57H870HrPVzO4xs2Ab6/4Q+CrwCy/fDG96e97rPzOztcAHbay7zX1lZgd7+/f/8+4HzOx57zVuefz5ZvYvL8P/mtlJrdb/ueM0OV+rZe83s3uTnjPA895zbpn+DzO7JukxX7TE35etlvgbcU3LfrTE+9KZ2ffNbIn3XJ43swPaeh1F0ppzTjfddGvjRuKbymuA24DfeNMuAB4HzgJWJi17D/A/JD44FgD3Ah8CQW/+fcAbwCCgN4liwwHHevNPBCpJfNMfAI4AtgL/5s0/FmjcQ9Y3gAf28nyOAWLAKUAW8DXv/pHe/HOBBuBXQA7wJSAOvA0cBQSB3wJLk9Z5v/eYs7x1ngDsAI7x5g8HpgL5QJ63L1cB2d7864BG4HJvm/nedAeUeb8/BPwJCHkZvgiUePOuBqLAeG/+dcB6oFdSvh1ehgDwLS/vgUnPoQr47l723d+AJ73fPwTuSJp3gPdaXe299jnACd68fGA5cKf3umd5+7Ko9fPc3etM4hjcBhwHmLe+ve3TU4DtwGne9nol7ctfA08lrb83UAcc1cbz/iKJD5whb9lHgTdbvf5t7l8Sx/MOEsd3FomiqgG4vx3vvxeB61pNa71/Wrb/NW/7FwE1JIrLod4+ehn4U9JjHgQWAwO91+pXQLhl/+0mx3XAi62mtee4ftH7/Xpv/SPb+V4/19tHv/H2+2gSZ6i+t7e/VR18rzcCc738+W2sd4/7CjgP+NSbfy2wBCjw5p1P4r35Je95nuq9NqPbcZyeC0RbZbkfuDfp/i7vndb7gcTxuhH4hbcfJ5B4L/7Mmz/CW8diYIC3/deTjxXddMukm+8BdNOtJ9/4rCg4yPvHlwW86/2D3VkUeP/wYsCJSY8tBOqBo5PmH580fwy7FgWLgV+22v4dLf8E2XtRsBS4cS/PZx7wYKtpDwP3eL+fS+IDaCBp/tvAnUn3J3q5e3v37wf+2WqdDwDz2shQ5D1+onf/OmDZbpZLLgruB/4OTNjNcp8A5yfdDwBrgelJj32q1WM2Aad34DgYTOJD2je8+zNJfIjL8+5fCbzTxmO/Q6JIyWpjfnuKgvl7ydd6nz4N3NTGssO943KId/9HJJqdtHdfHORtq6V42+P+JVHMLWw1/3U6tyhILnLyvXzfTpp2MfC/3u8DvPnDWx0z1bT6gJk0/zpaFQXtPK7/23sv/DfQL2nZvb3XzyXxPgwmzX8UmLuH7f+DXYuC9rzXdwChPayzXfvKew3+z8s8MWn6v4CzW63z73z2oX1Px+m57H9R8F0S/Swsaf6FQMT7fYS3ji8nzb+k5VjRTbdMu6n5kEg7OOf+ReJbwF8AJcCzrRYpJvFN1Iqkx9QAFcCwpPkrkx6zgl2VAld5p9GrLDHCyrkkPpC2xyYSZyn2ZNhutrvMm96iwjnXnHS/jsSH2uT7kPgQ1GJlq3WuJPEtLWY2wGuestoSzV/WeMsUJy2/ai+5f+bl/rslmuDcYYk+FJ97Tl72la2eU3J+gNpW+fdmBolvahd79x8g8e3qv3v3R5AoTnZnBLDcObc/I0etTL7Tjn3aZh7n3GrgBeAH3qQfkvjgvltmNsrMHvOa5Wwj8YE+eVuw5/07tHV+Pn8M7q+d23fO1bWeRuKYbclT6v38v6T32RYgm12PmT1q53E9gcQxcq1zbkvS9Pa81yucc01J9zt6zLbnvb7eORffwzrau69uAw4G/uacW9Lq8Xe2ep5f5bO/UyNo+33TGYYBq5xzLmla630Aux4rHd3PImlDRYFI+80jURTMb/XPGhIfyOMk/skBOzv+lpD4sLCZxLezI5Iek/w7JD4YX+ec65N0K3LOndrOfE8DU8ys7x6WWbOb7Y7ksw80+6r1OkeQ+LYe4AYSzWuOdM714rN/yJa0fHIR8jnOuU3OuZnOudHAZBLfFrcMO7jLc7JEh+AR7P9zSl7fDKAPsNbMNpBoIhEk8a0jJD70jmljFSuB0rbag5NoTlGQdH93RWDr/bO3fbqnPJBo6vYDMzuUxJmfhXtY9m4STTy+6G1rcqtt7c06dn98+KWlAB3T6r2W75x7uI3H7O74bM9x/SGJomCRmZ3QKsP+vNfboz3v9T2+72jHvjKzPGABiW/xv7Gb53leq8cWus/6OK2k7eN0O7u+L+Dz7w3Hnq0BDjSz5NekM/7eiaQlFQUi7fcwcBKJb8V24X07/V/A9WY22MzygVtItL192ysiHgJ+ZWYDzawXMKfVan4PzLLEOOxBM8sxs8PNbFI7891GoonTYjObZGbZZhYys6+Z2V3eMguAaWZ2sreNU0i0Af/PDu2JzzvKEp07g2Z2HImO2Au8eb1IfFO71SuUbuzoys3s382s1PvnXk2iwGopzO4HrjSzsWaWA/ycRDOvp/brGX1mCokPfMcAhyTdTiPxvA8mceZgnJld5XXMzEn6cPSUl3eumfW2RMfooyzRYR3gPeAc7zEjgMvakWlv+/RO4Edmdoq3vV6269CNT5E4c3UfsMg5t3Uv26oFqsxsAIk+CR2xEDjDzI73spwFHNnBdXQa51wFiffiXWY2BMDM+pjZN5POPrW2ARjuHV8t2nVcO+ceA6YDfzGz073J+/teb4/9fq+3c1/dSeJLjx+SaHrzYFJH3bnAdZbo9G1mlmdmZWY2PumxbR2nHwAlZnaaJTowfxP4t1YRN7Dn4rflOP8Pbx+PA64icdyLSCsqCkTaySVG8nlxDx+gZpHob/AOsJrEt4hTk84qXEridH4Y+IhE29qdZxycc8+T6Jh3E4l/sutJ/FNt64NK63zbgTISzTseIfHheTmJNuN/8ZZ5HTgHuJlEm/jfAWc5595qzzb24C8kOhFuJfEP9xJvWwC/JHHGpJJEu+M3SHre7XQo8CqJb9U/Bt4nsZ/wfj4MPE+iU+FxwEnOuW3tXbklRi9pa6SnC4HHnXPvOec2JN2eA94ELnTOfUri7MWJJM6QbCDR5AnnXK2XaRiJfh+bvcwtQ0D+mM86kv6FRJGzN3vcp865p0ic3fitt94IidGTWuY3kXidDmUPTYc8s0h0iN0G/JPPmlC1i3PuVeAnJDrebyFRZD3SkXV0gfNJ7JN/mNl2Eu/Hb9P2N8+Pkvh2eYPXDKaUDhzX3rFyOjDfzL63v+/19ujE93qb+8oSI6idSqKTfpNz7gESf9ceMrOgc+5P3nb/08uwmsTZ1mwvY5vHqXNuGYm/mfP47LhZ1Crbz4Ffmzc60272QTWJL3JOIPG34TkSX97c2sF9IJIRbNemdiIiHWOJK9M2Oud+6HcWaT9LDBt7tXNunN9ZRETEfzpTICKSYbymS5cCt/udRUREegYVBSIiGcTMfkqiKcUqEk0zRERE1HxIRERERCTT6UyBiIiIiEiGU1EgIiIiIpLhsvwO0NUCgYDLy8vzO4aIiIiIpLm6ujrnnEvJL93TvijIy8ujtrbW7xgiIiIikubMbIffGfZVSlYyIiIiIiLSeVQUiIiIiIhkOBUFIiIiIiIZTkWBiIiIiEiGU1EgIiIiIpLhVBSIiIiIiGQ4FQUiIiIiIhlORYGIiIiISIZTUSAiIiIikuFUFIiIiIiIZDgVBSIiIiIiGU5FgYiIiIhIhlNRICIiIiKS4VQUiIiIiIhkOBUFIiIiIiIZTkWBiIiIiEiGU1EgIiIiIpLhVBSIiIiIiGQ4FQUiIiIiIhlORYGIiIiISIZTUSAiIiIikuFUFIiIiIiIZDgVBSKyX7Y88CAbb7jB7xgiIiKyH1QUiMh+qfrrX9my4L+ofestv6OIiIjIPjLnnN8ZulRBQYGrra31O4ZIWnL19YQPnwQNDYTGjaP0sUVYMOh3LBERAZasmu13BF9MPHCOb9s2szrnXIFvAfaDzhSIyD6LL18ODQ0UnXgC8UiEqkWL/I4kIiIi+0BFgYjss1g4DEDxpZeSd9hhbLrtdppqanxOJSIiIh2lokBE9lk8HMFCIXJGjGDg1bNpqqyk8p55fscSERGRDlJRICL7LBYOExozBsvKIu/gg+l9+lS2LFhA/dp1fkcTERGRDlBRICL7xDlHPBwmd8L4ndOKZ82CQICKW272MZmIiIh0lIoCEdknjRUVNFVVERr3WVGQPWgQ/WfMYPszz1L3/vs+phMREZGOUFEgIvskVl4OQO74cbtM7z/jPLJKSth4wxxcc7Mf0URERKSDVBSIyD6JhyMAhMbtWhQE8vMpvmwWsY8+YtvixX5EExERkQ5SUSAi+yQWCZM9dCjBoqLPzes9dSq5Bx1Exa1zad6xw4d0IiIi0hEqCkRkn8TDEUKtmg61sECAgbOvonHDBirnz+/mZCIiItJRKgpEpMOa6+qoX7mS3KROxq3lT5pE0cknU3nvfTRs3NiN6URERKSjVBSISIfFly4F53YZjnR3Sq64HBob2TT3992UTERERPaFigIR6bBYSyfj8XsuCnKGDaPfOWdT/fjj7PjXx90RTURERPaBigIR6bBYuJxAYSHZQ4bsddn+F11EsF8/Ns65AedcN6QTERGRjlJRICId1tLJ2Mz2umywsJDimTPZ8e57bH/+hW5IJyIiIh2lokBEOsQ1NxOPRPbYybi1PmdMIzRmDBU33URzPN6F6URERGRfqCgQkQ5pWLuW5rq6Nocj3R3LyqJk9lU0rF3L1oULuzCdiIj47rVP4TfvQG2D30mkA1QUiEiHxMrDAOTupZNxa4WTJ1N47LFs/uPdNFZWdkU0ERHpCT7YDO9WQF6W30mkA1QUiEiHxCNhCAQIjRnT4ceWXHklzfE4m26/owuSiYhIjxCtgpG9IbD3fmfSc6goEJEOiYUj5JSWEsjN7fBjQyNL6Tt9OlWPPkos8kkXpBMREV81OVi+DUb39juJdJCKAhHpkFi4nNxx7e9P0FrxJRcTKCqi4sYbNUSpiEi6WVeDxZtUFKQgFQUi0m5N1dU0frqe0F6uZLwnwT59KL7kYmrfeIOaV1/txHQiIuK7aFXi5+g+/uaQDlNRICLtFoskrmTc0U7GrfWdPp2cESOouPF3uAaNTiEikjai1bicAAwv9DuJdJCKAhFpt3g4URSE9qP5EIBlZ1Ny5ZXUr1jB1j8/0hnRRESkJ4hWw4heENRHzFSjV0xE2i0WDhPs14+s4uL9XlfhV48l/+ij2PyHP9BUXd0J6URExFfOwbJqNR1KUSoKRKTd4uEwuePHY7b/w8yZGQNnz6Zp+3Y233VXJ6QTERFfbazDahpgjDoZpyIVBSLSLq6hgXg0Smg/+xMkyx03jj7TprHlwYeIr1jRaesVEREfLPXO+o5SUZCKVBSISLvEV6zA1deTO37/+hO0VnzpTAKhEBU33dyp6xURkW62rBoXMCjt5XcS2QcqCkSkXeLeyEOdeaYAIGvAAPpfeCE1L79M7Ztvduq6RUSkG0Wr4MAiyAn6nUT2gYoCEWmXWDiMZWcTKi3t9HX3O+dssocMYeOcG3FNTZ2+fhER6QbRal20LIWpKBCRdomHI+SMGY1lZ3f6ugOhECVXXE48EqFq0aJOX7+IiHSxyhi2Na6iIIWpKBCRdomFw+SO69ymQ8mKpkwh77DD2HTb7TTV1HTZdkREpAu0XMl4lIYjTVUqCkRkrxo3baKpspLcCV1XFJgZA6+eTVNlJZX3zOuy7YiISBeItow8pE7GqUpFgYjsVWznlYy7rigAyDv4YHqfPpUtCxZQv3Zdl25LREQ60bJq3NACyO/8JqbpxsymmFnEzKJmNns384eb2Stm9r9m9n9mdmp35FJRICJ7FQuXA3T6cKS7UzxrFgQCVNyiIUpFRFJGtEpNh9rBzILAncApwERguplNbLXYNcBfnHOHAmcC3XKFTxUFIrJX8XCErMEHEOzd9R3IsgcNov+MGWx/5lnq3n+/y7cnIiL7aVs9tnGHOhm3zxFA1Dm33DlXD/wZOL3VMg5oaYfVG/i0O4KpKBCRvYpFuraTcWv9Z5xHVkkJG2+Yg2tu7rbtiojIPljm9SdQUdAeQ4A1SffXetOSXQecZWZrgaeBn3RHMBUFIrJHzbEY9StWEuqGpkMtAvn5FF82i9hHH7Ft8eJu266IiOyDlpGHRqv5UN9gMKt8/IR3k24X7MNqpgP3O+eGAqcCC82syz+zZ3X1BkQktcWXRqGpqVvPFAD0njqVrQ88SMWtcyk68UQCeXndun0REWmnaDWuJA965fidxHdbm5oaJ4TLJ+1hkXXAsKT7Q71pyWYAUwCcc2+aWS4wAKjozKyt6UyBiOxRPBIG6NLhSHfHAgEGzr6Kxg0bqJw/v1u3LSIiHaArGXfEO8AYMys1sxwSHYmfbLXMauB4ADObAOQCm7o6mIoCEdmjWDhCID+f7GHD9r5wJ8ufNImik0+m8t77aNi4sdu3LyIie7GjEdbVaOShdnLONQI/Bp4DykmMMvSxmf3azKZ6i10OnG9mHwIPA+c651xXZ1NRICJ7FAuXExo7Fgv48+ei5IrLobGRTXN/78v2RURkD5ZXYw6dKegA59zTzrmxzrlRzrnfeNN+6Zx70vt9iXNusnPuS865Q5xzz3dHLhUFItIm5xzxyCeEurnpULKcYcPod87ZVD/+ODv+9bFvOUREZDeiGnkoXagoEJE2Naz7lObt27u9k3Fr/S+6iGC/fmy84Qa64QyqiIi019IqXJ8Q9M/1O4nsJxUFItKmnZ2Mu3E40t0JFhZSPHMmO957j+3PdctZVBERaY9lXidjM7+TyH5SUSAibYqVh8GM0NixfkehzxnTCI0ZQ8XNN9Mcj/sdR0RE6ptg1XYYpaZD6UBFgYi0KR4Jk3PggQTy8/2OgmVlUTL7KhrWrmXrwoV+xxERkVXbsSYHYzTyUDpQUSAibYqFI4TG+9ufIFnh5MkUHnssm/94N42VlX7HERHJbEtbrmSsMwXpQEWBiOxWU00NDWvW+N6foLWSK6+kOR5n0+13+B1FRCSzRatxBVkwyP+zybL/VBSIyG7FIxGAHnWmACA0spS+06dT9eijxCKf+B1HRCRzLatOXLRMnYzTgooCEdmtWLhl5KGeVRQAFF9yMYGiIipuvFFDlIqI+KGpGZZXq+lQGumWosDM5ptZhZn9K2laPzN7wcyWej/7etPNzG43s6iZ/Z+ZHZb0mHO85Zea2TndkV0kU8XDYYK9e5M1cKDfUT4n2KcPxZdcTO0bb1Dz6qt+xxERyTxrarD6ZhUFaaS7zhTcD0xpNW028JJzbgzwkncf4BRgjHe7APgjJIoI4FrgSOAI4NqWQkJEOl9LJ2ProaeF+06fTs6IEVTc+DtcQ4PfcUREMouuZJx2uqUocM79N7Cl1eTTgQXe7wuAbyRN/y+X8BbQx8wOAE4GXnDObXHObQVe4POFhoh0AtfURHzp0h7ZdKiFZWdTcuWV1K9YwdY/P+J3HBGRzBKtwoWCMLTI7yTSSfzsUzDQObfe+30D0NJGYQiwJmm5td60tqaLSCerX7UKF4v1uE7GrRV+9Vjyjz6KzX/4A03V1X7HERHJHNFqGNkLgj3zbLJ0XI/oaOwSPQU7rbegmV1gZu+a2buNjY2dtVqRjBErLwfoccORtmZmDJw9m6bt29l8111+xxERyQzNLjHykJoOpRU/i4KNXrMgvJ8V3vR1wLCk5YZ609qa/jnOuXnOuUnOuUlZWVmdHlwk3cXDEcjOJjRqlN9R9ip33Dj6TJvGlgcfIr5ihd9xRETS3/parK4xMRyppA0/i4IngZYRhM4BnkiafrY3CtFRQLXXzOg54CQz6+t1MD7JmyYinSwWCRMaORLLyfE7SrsUXzqTQChExU03+x1FRCT9LVMn43TUXUOSPgy8CYwzs7VmNgOYA5xoZkuBE7z7AE8Dy4Eo8CfgYgDn3BbgeuAd7/Zrb5qIdLJ4ebjHNx1KljVgAP0vvJCal1+m9s03/Y4jIpLeotW4LIMD1ck4nXRL2xrn3PQ2Zh2/m2UdcEkb65kPzO/EaCLSSuOWLTRu2kRo/AS/o3RIv3POpuqRR9g450ZKH1uEBYN+RxIRSU/RKjiwF+To72w66REdjUWk54jvvJJx6pwpAAiEQpRccTnxSISqRYv8jiMikp6cS4w8pKZDaUdFgYjsIhaOAPT44Uh3p2jKFPIOO4xNt91OU02N33FERNLP5hhWXa+iIA2pKBCRXcTC5WSVlJDVN/UuGG5mDLx6Nk2VlVTeM8/vOCIi6SdalfipkYfSjooCEdlFPBwhNCH1zhK0yDv4YHqfPpUtCxZQv3a3oxaLiMi+WlaNM2BUL7+TSCdTUSAiOzXX1xNfvpzccalbFAAUz5oFgQAVt2iIUhGRTrW0GoYVQq6uA5VuVBSIyE710Sg0NqZcJ+PWsgcNov+MGWx/5lnq3n/f7zgiIukjWqWmQ2lKRYGI7PRZJ+PUGo50d/rPOI+skhI23jAH19zsdxwRkdRXFcc2x9TJOE2pKBCRneKRMJabS86Bw/2Ost8C+fkUXzaL2EcfsW3xYr/jiIikPl3JOK2pKBCRnWLhCKGxY9Pmwl+9p04l96CDqLh1Ls07dvgdR0QktUW9omCUioJ0pKJARABwzhELh8kdl9r9CZJZIMDA2VfRuGEDlfN1MXQRkf0SrcINyoeiHL+TSBdQUSAiADRu2EBzdXVKD0e6O/mTJlF08slU3nsfDRs3+h1HRCR1Rat1liCNqSgQEQBi4TAAuSl4JeO9KbnicmhsZNPc3/sdRUQkNdU2YJ/Wqj9BGlNRICIAxL2iIDQ2fZoPtcgZNox+55xN9eOPs+Ojf/kdR0Qk9Sz3+hOM0XCk6UpFgYgAiU7G2cOHEyws8DtKl+h/0UUE+/Vj45w5OOf8jiMiklqWqpNxulNRICJA4kxBOnUybi1YWEjxzJnseO89tj/3vN9xRERSy7JqXL8Q9Mv1O4l0ERUFIkJzbS31q1cTSvErGe9NnzOmERozhoqbb6Y5Hvc7johI6ohWwWg1HUpnKgpEhNgnn4BzadnJOJllZVEy+yoa1q5l68KFfscREUkN8SZYXaNOxmlORYGIEI9EgPQceai1wsmTKTz2WDb/8W4aKyv9jiMi0vOt2IY1OxUFaU5FgYgQC4cJ9OpF1uDBfkfpFiVXXklzPM6m2+/wO4qISM8XrUr8HKXmQ+lMRYGIEC8Pkzt2LGbmd5RuERpZSt/p06l69FFikU/8jiMi0rMtq8YVZsPAPL+TSBdSUSCS4VxzM7GlSwlNmOB3lG5VfMnFBIqKqLjxRg1RKiKyJ9HqRNOhDPniKFOpKBDJcA2rV+Pq6shN85GHWgv26UPxJRdT+8Yb1Lz6qt9xRER6psZmWLFNIw9lABUFIhkuFk50Mg6NS/9Oxq31nT6dnBEjqLjxd7iGBr/jiIj0PKu3Yw3N6mScAVQUiGS4WLgcgkFCY0b7HaXbWXY2JVdeSf2KFWz98yN+xxER6Xmi3pWMVRSkPRUFIhkuHo4QGllKIBTyO4ovCr96LPlHH8XmP/yBpupqv+OIiPQs0SpcbhCGFPqdRLqYigKRDBeLRDKy6VALM2Pg7Nk0bd/O5rvu8juOiEjPEq2GUb0hoE7G6U5FgUgGa6qqonH9+ozrZNxa7rhx9Jk2jS0PPkR8+Qq/44iI9AzNDpZVq+lQhlBRIJLBdnYyHp9Zw5HuTvGlMwmEQlTcdJPfUUREeoZ1NVisKXGmQNKeigKRDBaPhAEy/kwBQNaAAfS/8EJqXnmF2jff9DuOiIj/lrV0MtZwpJlARYFIBouFIwQHDCBrwAC/o/QI/c45m+whQ9g450ZcU5PfcURE/BWtxj/dImwAACAASURBVGUH4MAiv5NIN1BRIJLBYuEwueN0lqBFIBSi5IrLiUciVC1a5HccERF/RathRBFk6eNiJtCrLJKhXEMD9dEouRMyd+Sh3SmaMoW8ww5j022301RT43ccERF/OAfRKjUdyiAqCkQyVHz5ClxDQ0YPR7o7ZsbAq2fTVFlJ5T3z/I4jIuKPih3Y9gZ1Ms4gKgpEMlQ8XA6ok/Hu5B18ML1Pn8qWBQuoX7vO7zgiIt2vpZPxGBUFmUJFgUiGioUjWE4OOaWlfkfpkYpnzYJAgIpbbvY7iohI91tahQsYlKooyBQqCkQyVDwSJjRmDJaV5XeUHil70CD6z5jB9meepe799/2OIyLSvaLVMKwQQkG/k0g3UVEgkoGcc8TCEUJqOrRH/WecR1ZJCRtvmINrbvY7johI99GVjDOOigKRDNRYsYmmLVvIVSfjPQrk51N82SxiH33EtsWL/Y4jgKuvp7m21u8YIultSwyrjGnkoQyjokAkA+28krGGI92r3lOnknvQQVTcOpfmHTv8jpOxmmpqqLz3XpYedzyfHH0MG2+YQ2Nlpd+xRNLTzisZ60xBJlFRIJKBYuEIACFduGyvLBBg4OyraNywgcr58/2Ok3Eat2yh4rbbiB53PBU330Lu2DH0OuUUtixcSPSEE6m45Vaaqqr8jimSXqJeUaDhSDOKigKRDBQPl5M9eDDBXr38jpIS8idNoujkk6m89z4aNm70O05GaFi/ng2//S3R446n8o93U3DkkYx49C8Mnz+fwTfOYeTixRQddxyV995L9IQT2XTHH2javt3v2CLpIVqFO6AACrL9TpKWzGyKmUXMLGpms3czf66ZfeDdPjGzbvnmQ0WBSAaKhSOEJkzwO0ZKKbnicmhsZNOtc/2OktbiK1bw6c9/TvSkk9n64EP0mjKFkU8tZugdt5N38ME7lwuNLGXILTdT+sTjFBx9NJvvvJPoCSeyed6f1OdAZH+pk3GXMbMgcCdwCjARmG5mE5OXcc7Ncs4d4pw7BLgDeKw7sqkoEMkwzTt2UL9yJblqOtQhOcOG0e+cs6l+4gl2fPQvv+OkndiSJaz96SyWn/o1ti1+ir7f+Q6jn3+OwXNuIDRqVJuPyx07lqF33M6IRX8l/5BD2HTrrURPPInK+++nORbrxmcgkia212Pr63TRsq5zBBB1zi13ztUDfwZO38Py04GHuyOYigKRDBNfuhSamzUc6T7of9FFBPv1Y+OcOTjn/I6TFurefZfV51/Aim9No/a11+h//vmMfulFBv3iGrKHDGn3evK+8AWG3XM3Bz78ELnjx1Ex50aWnXQyWx56CFdf34XPQCTNtHQyHqWRh7rIEGBN0v213rTPMbMDgVLg5W7IpaJAJNPEwi0jD6n5UEcFCwspnjmTHe+9x/bnnvc7TspyzrH9H/9g5Xe/x6qzvk/s448pnjWL0a+8TMlls8gaMGCf151/6KEMnz+f4QsWkD1sGBt/fT3LppxC1aJFuMbGTnwWImlKIw/tl77BYFb5+AnvJt0u2I/VnQn81TnX1Fn59kSXMhXJMPFwhEBBQYe+hZXP9DljGlsffJCKm2+m8KvHEgiF/I6UMlxTE9uefZbKP91LPBwma/ABDLzmGvpM+xaBvLxO3VbBkUeQ/8BCal9/g0233cb6n1/D5nnzKP7xj+l16qlYUFdpFdmtaDVuQC700d+2fbG1qalxQrh80h4WWQcMS7o/1Ju2O2cCl3RWtr3RmQKRDBOLRAiNG4cF9PbfF5aVRcnsq2hYu5atCxf6HSclNNfXs/Uvf2HZqafy6eVX4OrrOeCGGxj93HP0O+t7nV4QtDAzCssmM+IvjzD0rjsJ5OXz6c+uZPnpp7Pt2ed0lWqR3YlW6aJlXesdYIyZlZpZDokP/k+2XsjMxgN9gTe7K5g+FYhkENfcTDwcJlf9CfZL4eTJFB57LJv/eLcuoLUHzbW1VP7n/Sw74UQ2/PJagoVFDLn9NkYu/jt9vvkNLLt7hjs0M4qOO47SxxYx5Pdzodmx7qc/ZcW0M9j+8ivqHyLSItYIa2vUdKgLOecagR8DzwHlwF+ccx+b2a/NbGrSomcCf3bd+AdKzYdEMkjDunU019YSGq8rGe+vkiuvZPnUqWy6/Q4O+NV1fsfpUZqqqtjywINsXbiQpupq8o84ggNu+C0FxxyDmfmWywIBek2ZQtGJJ7LtqafY9Ic7WXvxxeR+8YsUXzrT93wivlu+DWsGp6KgSznnngaebjXtl63uX9edmUBFgUhG2dnJWEXBfguNLKXv9OlsffBB+n73u+SOG+t3JN81bKxgy/33U/XIIzTX1VH41a/S/4LzyT/0UL+j7cKCQXpPnUqvU06h+okn2HTXXayZ8UPyJ02i+NKZ5H/5y35HFPFH1LtGlkYeykhqPiSSQeLlYQgECI0Z43eUtFB8ycUEioqouPHGjG6CUr96Net/eS3LTjiBLQsWUHjccZQ+8QTD/nhXjysIkll2Nn3OOINRzz7LwF9cQ/2qVaz6/tmsPm8GOz780O94It1vWTWudw4U5/qdRHygokAkg8QiEXJGjOiyjp2ZJtinD8WXXEztG29Q8+qrfsfpdrFIhHWXX8GyKadQ/be/0ftb32LUs88w5OabUurMSSAnh37f+x6jXniekquuIlZezsp/P5M1F/2I2JIlfscT6T7RahjVG9SMLiOpKBDJIOpk3Pn6Tp9OzogRVNz4O1xDg99xukXd//4vay76EStO/wY1r7xCvx+cy6iXXuSAX11HzvDhfsfbZ4HcXPr/4FxGv/gCxT/9KXXvv8+Kb01j7cxLExf9E0lnDc2wcps6GWcwFQUiGaJp2zYa1q0jNE79CTqTZWdTcuWV1K9YwdY/P+J3nC7jnKPmtddZ9f2zWTX9u+z44AMGzPwJo19+iYE/+xnZJSV+R+w0gYICBlx0IaNffIEBF19M7euvs3zq6ay74mfUr1zpdzyRrrFqG9boNBxpBlNHY5EMEY9EAHSmoAsUfvVY8o8+is1/+AO9v34awT7p80/VNTez/YUXqZw3j9jHH5M1cCADr55Nn29/m0B+vt/xulSwVy+KZ/6Evt8/iy3z57PlgQfZ9swz9P7G6Qz40cXkDNUFACWNRHUl40ynMwUiGSIWThQFofETfE6SfsyMgbNn07R9O5vuusvvOJ3CNTRQ9djfWP6101h36aU01WzngP93PaNeeJ5+55yT9gVBsqy+fSm5/HJGv/A8/c76Htv+vphlp5zC+l/9ioaNG/2OJ9I5otW4/Cw4oMDvJOITFQUiGSIWCRPs25eskmK/o6Sl3HHj6DNtGlsfepj48hV+x9lnzTt2sGXhA0RPOpn1//EfWCjEkLm3Murpp+lzxhkEcnL8juibrAEDGHj11Yx6/jn6nDGNqr8uYtmJJ7Hxhhto3LzZ73gi+ydalehkHFAn40ylokAkQ8TLw4TGj9PFmbpQ8aUzCYRCVNx0k99ROqxp2zY2330P0eNPYONvfkP24MEMm3cPpX97jF6nnIIFg35H7DGyBw3igGuvZdQzT9PrtNMSRdSJJ1Fxy600bt3qdzyRjmtysHxboiiQjKWiQCQDuMZG4kuXkqumQ10qa8AA+l94ITWvvELtm2/6HaddGjdvpuKWW4kedzybfv97cg/6Agc+sJARDz5A4b/9m4rIPcgZOpTBv/0NI59aTNHxx1N5770sO+FENt3xB5q2b/c7nkj7ravB4k3qT5DhVBSIZID6lStx9fXqZNwN+p1zNtlDhrBxzo24pia/47Spfu06Nvz6eqLHn0DlvfdS8JUySh9bxPB588ifNMnveCklVFrKkJtvovSJxymYPJnNd95J9IQT2XzPPJpra/2OJ7J3LVcyHpM+gyRIx6koEMkAn3Uy1nCkXS0QClFyxeXEIxGqFi3yO87nxKNRPr1qNstOPpmtjz5K76lfZ9QzTzN07lxyJ070O15Kyx07lqG338aIRX8l/5BD2DR3LtETT6Ly/vtpjsX8jifStqXVuJwADCv0O4n4SEWBSAaIh8shO5tQaanfUTJC0ZQp5B12GJtuu52mmhq/4wCw46OPWPuTn7D8tK+z7fnEKDqjX3ieA66/npwRI/yOl1byvvAFht1zNyP+/DC548dRMedGlp10Mlseeojm+nq/44l83rJqKO0FQX0szGR69UUyQCwcITR6NJbBI8d0JzNj4NWzaaqspPKeeb7lcM5R+9ZbrD7vPFZ++zvU/s/bDLj44sQFx66+muxBg3zLlgnyDjmE4fPnM/y/FpA9fBgbf309y6ecQtVf/5oxV7+WFOBcoijQRcsynooCkQwQi4TJHaf+BN0p7+CD6X36VLYsWED92nXdum3X3Mz2l15i5ZlnsvrcHxBbupSSn13B6JdfpnjmT8jq27db82S6giOO4MCFCxl2770EBwxg/TW/YNlpp1H95JM9ut+JZIiNdVhNgzoZi4oCkXTXuHkzTZs2E1In425XPGsWBAJU3HJzt2zPNTZS/eSTrDj9dNZe8mOaKrcw6LprGf3ii/SfMYNgoS5K5Bczo7BsMiMe+TND77qLQF4+n155Fcunns62Z5/DNTf7HVEy1VJdyVgSVBSIpLmWTsYajrT7ZQ8aRP8ZM9j+zLPUvf9+l22nOR5n68MPs2zKKXx65VWAMfimmxj17DP0PfNMAqFQl21bOsbMKDruq5Q+toghv58LwLqf/pQV35rG9pdfwTnnc0LJONEqXMBgRC+/k4jPVBSIpLl4JAyg4Uh90n/GeWSVlLDxhjmd/m1wU00NlffeS/SEE9jwq1+T1b8/Q++6k9InHqf310/DsrI6dXvSeSwQoNeUKYx88gkG/+5GmuvqWHvxxaz89zOpee11FQfSfZZVw4FFkKMLFGY6FQUiaS4WjpA1aBDBPupE5odAfj7Fl80i9tFHbFu8uFPW2bh1KxW33Ub0uOOpuPkWcseMZfiCBRz454cpOu44LKA/7anCgkF6T53KqKcWc8D/u57GzZtY88Mfsur736funXf8jifpzrlE8yE1HRJUFIikvXi4XJ2MfdZ76lRyDzqIiltupbmubp/X07B+PRt++1uixx1P5d33UHDkkYx49FGGz7+PgiOP0NWHU5hlZ9PnjDMY9eyzDPzlL2hYtZpV3z+b1eedx44PPvA7nqSrLTGsKq6iQAAVBSJprTkeJ758BaEJumiZnywQYODsq2jcuJHK+f/Z4cfHV6zg05//nOhJJ7P1oYfpdfLJjFz8d4becTt5Bx/UBYnFL4GcHPp997uMeuF5SmZfRSwcYeWZ01lz4UXElizxO56km2hLJ2OdSRZQg1ORNBaPRqGpiVxdydh3+ZMmUXTyyVTedx99vn0G2QMH7vUxsSVL2DzvT2x/7jksJ4e+3/kO/c/7AdlDhnRDYvFTIDeX/ueeS99vf5stDzxI5fz5rPjWNIpOOonin/yY0JgxfkeUdBCtxhkwUp2MRWcKRNJaPJzoZBxS86EeoeSKy6GxkU23zt3jcnXvvsvq8y9gxbemUfvaa/Q//3xGv/wSg35xjQqCDBMoKGDAhRcw+sUXGHDJJdS+/jrLp57Ouit+Rv3KlX7Hk1QXrYIhBZCf7XcS6QFUFIiksVg4guXnkzN8uN9RBMgZNox+55xN9RNPsOOjf+0yzzlHzauvsvK732PVWd8n9vHHFM+axehXXqbksllk9e/vU2rpCYJFRRT/5MeMevEF+v/wh2x/6SWWfe00Pv35z7v94niSRnQlY0miokAkjcXDYXLHjMGCGmqup+h/0UUE+/Vj45w5OOdwTU1UP/UUK775LdZceBENG9Yz8JprGP3Siwy48AKCRUV+R5YeJKtvX0ouv4zRLzxPv7O+x7a/L2bZKaew/le/omHjRr/jSSrZVo9t3KFOxrKTigKRNOWcIxaJEFJ/gh4lWFhI8cyZ7HjvPTZefz3LTj2VTy+/AldfzwE33MDo556j31nfI5CX53dU6cGyBgxg4NVXM+qF5+lzxjSq/rqIZSeexMYbbqBx82a/40kqWOZ1Mh6lokASVBSIpKnGTz+leds2XbSsB+pzxjRCY8ey9aGHCRb1YsgdtzNy8d/p881vYNlq2yvtlz1wIAdcey2jnnmGXl8/jS0PPEj0xJOouOUWGrdu9Tue9GTRqsRPNR8Sj0YfEklTsUgEQGcKeiDLymLY3X+kYf168g47TNcXkP2WM3QIg3/zG/r/8IdsvvMuKu+9j60PPUy/c86h3w/OVTM0+bxoNa4kD3rl+J1EegidKRBJU7FwGMzIHTvW7yiyG9mDB5N/+OEqCKRThUpLGXLzTYx88gkKJk9m8113ET3hRDbfM4/m2lq/40lPEtWVjGVXKgpE0lS8PEz28GEECgr8jiIi3Sw0ZgxDb7+N0scWkX/ooWyaO5eVZ07HOed3NOkJ6hpgXY2aDskuVBSIpKlYJELu+Al+xxARH+VOnMiwu//IwJ//nPjSpcQ/+cTvSNITLN+GOdTJWHahokAkDTXV1NKwerU6GYsIAEUnnQRA7Wuv+ZxEeoSWkYfGqCiQz6goEElDLd8Ghsapk7GIQPbAEkJjx1KjokAAllbh+oagX67fSaQHUVEgkoZi4XIAcieoKBCRhIKyMna8+x7NdXV+RxG/RasTTYc00IEk8b0oMLNZZvaxmf3LzB42s1wzKzWz/zGzqJk9YmY53rIh737Umz/C3/QiPVM8HCHQuzdZgwb5HUVEeojCssm4hgZq337b7yjip/omWL1dIw/J5/haFJjZEGAmMMk5dxAQBM4EbgTmOudGA1uBGd5DZgBbvelzveVEpJVYJEzuuHEa7lJEdso7/HAsN5fa1173O4r4aeU2rMlp5CH5HN/PFJC4gFqemWUB+cB64Djgr978BcA3vN9P9+7jzT/e9KlHZBeuqYl45BNC6mQsIkkCoRD5R3xZnY0zXdTrZKwzBdKKr0WBc24dcDOwmkQxUA28B1Q55xq9xdYCQ7zfhwBrvMc2esv3787MIj1d/arVuFhMw5GKyOcUlpVRv3Il9WvX+R1F/BKtxhVkwaB8v5NID+N386G+JL79LwUGAwXAlE5Y7wVm9q6ZvdvY2Lj3B4ikkXgkDKDhSEXkcwrKvgJoaNKMFq2CUX3UyVg+x+/mQycAK5xzm5xzDcBjwGSgj9ecCGAo0PKVxjpgGIA3vzdQ2Xqlzrl5zrlJzrlJWVlZrWeLpLVYOAJZWeSMHu13FBHpYXJKR5A9eDC1r6soyEhNzbBim5oOyW75XRSsBo4ys3yvb8DxwBLgFeAMb5lzgCe835/07uPNf9npmu0iu4iFywmNHEkgJ8fvKCLSw5gZBWVl1L7xJq6hwe840t3W1GD1zbpomeyW330K/odEh+H3gY+8PPOAq4DLzCxKos/Afd5D7gP6e9MvA2Z3e2iRHi4ejqiTsYi0qaBsMs21tez48EO/o0h3W1qV+DlKRYF8nu9ta5xz1wLXtpq8HDhiN8vGgG93Ry6RVNS4dSuNGzeSqysZi0gbCo4+GoJBal57jfxJk/yOI91pWTUuFIShRX4nkR7I7+ZDItKJ4uFEJ2OdKRCRtgSLisj70pd0vYJMFK2Gkb0gqE7G8nkqCkTSSCwcASB3vM4UiEjbCsomE/v4Yxq3bvU7inSXZgfLqtXJWNqkokAkjcTDYbKKi8nqr8t3iEjbCr/yFXCO2tff8DuKdJf1tVhdo65kLG1SUSCSRmLhMCGdJRCRvcidOJFgnz66XkEm0ZWMZS9UFIikCVdfT3z5cjUdEpG9smCQgmOOoeb119DI3hliWTUuy+DAXn4nkR5KRYFImogvXw4NDepkLCLtUlBWRtOmzcQjEb+jSHeIVsGIXpCtj36yezoyRNJEzBt5SGcKRKQ9CiZPBlATokzgXKL5kK5PIHugokAkTcTLw1goRM6BB/odRURSQPbAEkJjx1KjoUnT3+YYVl2v/gSyRyoKRNJELBIhNHYsluX7NQlFJEUUlJWx4733aK6r8zuKdKWodyVjjTwke6CiQCQNOOeIh8Pkqj+BiHRAYdlkXEMDtW+/7XcU6UrRapyRuHCZSBtUFIikgcaNG2mqqiI0Tv0JRKT98g4/HMvL09WN0120GoYVQq7OJEvbVBSIpIGdnYwnqCgQkfYLhELkH/Flav/5T7+jSFeKVqnpkOyVigKRNBAPJ4YUDI1T8yER6ZjCyWXUr1pF/dq1fkeRrlAVxzbHNPKQ7JWKApE0EIuEyR46lGBhod9RRCTFFJSVARqaNG0t05WMpX1UFIikgXh5WBctE5F9klM6guzBg6lRUZCeNPKQtJOKApEU11xXR/2qVeSOn+B3FBFJQWZGQVkZdW++hWto8DuOdLZoNW5QPhRm+51EPGY2xcwiZhY1s9ltLPMdM1tiZh+b2UPdkUtFgUiKiy9dCs5pOFIR2WcFZZNprq1lx4cf+h1FOlu0Wk2HehAzCwJ3AqcAE4HpZjax1TJjgKuByc65LwA/7Y5sKgpEUlysPDHyUGi8Rh4SkX1TcPTREAyqCVG6qW3APq1V06Ge5Qgg6pxb7pyrB/4MnN5qmfOBO51zWwGccxXdEUxFgUiKi0XCBAoLyR4yxO8oIpKigkVF5B1yCLX/VFGQVpark3EPNARYk3R/rTct2VhgrJm9bmZvmdmU7gimq1iIpLh4OEJo/DjMzO8oIpLCCssms+n2O2jcsoWsfv38jiOdYamKgu7WNxjMKh8/4d2kSfMmhMvndXA1WcAY4FhgKPDfZnawc66qk2K2uVERSVGuuZl4JELvb37T7ygikuIKysrYdNvt1L7+Br2/fprfcaQzRKtw/ULQN9fvJBlja1NT44Rw+aQ9LLIOGJZ0f6g3Ldla4H+ccw3ACjP7hESR8E6nhm1FzYdEUljDmjU019VpOFIR2W+5EycS7NNH1ytIJ8uq1Z+g53kHGGNmpWaWA5wJPNlqmcdJnCXAzAaQaE60vKuDqSgQSWEx70rGGo5URPaXBYMUHHMMNW+8jnPO7ziyn5p37IDV29V0qIdxzjUCPwaeA8qBvzjnPjazX5vZVG+x54BKM1sCvAL8zDlX2dXZ1HxIJIXFI2EIBAiNGe13FBFJAwVlZWx7+mnikQi5GtEspcU/+QRrBqeioMdxzj0NPN1q2i+TfnfAZd6t2+hMgUgKi5WHySktJZCr9qIisv8KyiYDqAlRGoiVlyd+UfMhaScVBSIpLBYJ69s8Eek02SUlhMaNo0ZDk6a82MdLcEXZUJLndxRJESoKRFJUU3U1jZ+uVydjEelUBWWTqXv/fZpra/2OIvshVl4Oo3qDhquWdlJRIJKiYpGWTsY6UyAinaewrAwaGqh9+22/o8g+cg0NxCMRGKOmQ9J+KgpEUlQ8HAYgNE5nCkSk8+QdfjiWl0fta6/7HUX2UXzZMlxDg0Yekg5RUSCSomLhCMH+/ckqLvY7ioikkUBODvlHfFmdjVNYbInXyXiUigJpPxUFIikqHg6TO24cpvaiItLJCieXUb9qFfVr1/odRfZBbMkSAvn5MKTQ7yiSQlQUiKQg19BAfOlSQupPICJdoOArZYCGJk1VsSVLEv8fAvrSSNpPRYFICoqvWIFraCB3gooCEel8OSNGkD1kiIYmTUGuuZlYOEzuxIl+R5EUo6JAJAXFvZGH1MlYRLqCmVFQVkbdW28lOqxKyqhfuQpXV6eiQDpMRYFICoqFw1h2NqHSUr+jiEiaKiibTHNtLTs++MDvKNIBsSVLAMidOMHnJJJqVBSIpKB4eZjQmDFYdrbfUUQkTRUcdRQEg9RoaNKUEitfkvjSaNQov6NIilFRIJKCYpGIOhmLSJcKFhWRd8gh6mycYmJLlhAaN05fGkmHqSgQSTGNmzbRVFlJ7nj1JxCRrlVYNpnYkiU0btnidxRpB+cc8SXl5E5Q0yHpOBUFIikmtvNKxjpTICJdq6DsK+Acta+/4XcUaYfGTz+lqbqa3C+ok7F0nIoCkRTTUhToTIGIdLXcL0wk2Lcvta/90+8o0g47dnYyVlEgHaeiQCTFxMMRsgYfQLC3Ll8vIl3LAgEKjjmGmtffwDU3+x1H9iJeXg7BIKGxY/2OIilIRYFIiolFwuSq6ZCIdJOCsjKaNm/eeX0U6bliHy8hNHIkgdxcv6NIClJRIJJCmmMx6pev0JWMRaTbFEw+BoAajULU48WWLNH1CWSfqSgQSSHxpVFoblYnYxHpNtklJYTGjaNW1yvo0Ro3baJx0yb1J5B9pqJAJIXEI+pkLCLdr6BsMnXvv09zba3fUaQNsfJyQJ2MZd+pKBBJIbHyMIH8fLKHDfM7iohkkMKyMmhooPbtt/2OIm2IeSMPhXSNAtlHKgpEUkgsEk5cqTKgt66IdJ+8ww/H8vKo/af6FfRUsSXlZB84nGBhod9RJEXpk4VIinDOEY98QkhNh0SkmwVycig44ghqXldR0FMlOhmr6ZDsOxUFIimiYd2nNG/fruFIRcQXBWVlNKxaTf2aNX5HkVaaqqtpWLuW3AkqCmTfqSgQSRHxsNeJTMORiogPCsomA1CroUl7nFi5NwiFzhTIflBRIJIiYuEImBEaM8bvKCKSgXJGjCB7yBBqNDRpj9PSyVjXKJD9oaJAJEXEI2FyDjyQQH6+31FEJAOZGQVlZdS99RauocHvOJIkVl5O1qBBZPXr53cUSWEqCkRSRKw8TGi8mg6JiH8KyibTXFvLjg8+8DuKJFEnY+kMKgpEUkBTTU2iE5mKAhHxUcFRR0FWFjUamrTHaK6ro375cnJ1fQLZTyoKRFJAPBIB0HCkIuKrYFEReYd8SZ2Ne5BYJALOkfsFnSmQ/aOiQCQF7BxZQmcKRMRnhWVlxJYsobGy0u8oQnInYxUFsn9UFIikgHgkTLBPH7IGDvQ7iohkuILJZQDUvvGGz0kEEkVBsG9f/X+Q/aaiQCQFxMIRQuPHY2Z+RxGRDchRdAAAIABJREFUDJf7hYkE+/ZVE6IeIlZeTu7Eifr/IPstqz0LHbzg4AJgJjAJKEqe99E5H53UBblExOMaG4l/8gl9zzzT7ygiIlggQMExx1Dz+hu45mYsoO8X/eLq64kvjVJ47rl+R5E00K6iAJgPHAo8DtR2XRwRaa1+1SpcPK7hSEWkxygoK2PbU08Rj0Q06o2P4tEoNDToomXSKdpbFJwEjP3onI82dWUYEfm8WNjrZDxBRYGI9AwFk48BoOafr6ko8JE6GUtnau85v0qgpiuDiMjuxcMRyM4mNHKk31FERP5/9u49PM66zv//653DzDQTCj1y6JFKaaYYKIilS6aux/3BHsD96tf1XJSDJzzvIovredcFUfGwKrKABoRFv7qu6CJeuypKyhkpxs4909aWHoGWUkqT9J5kZj6/P3KnhjZNZpLJ3DOT5+O6cnXue+7c96t3J+m853OSJDXPnatoWxvjCkLmp1JqaG1V84IFYUdBHSi2KLhK0tfaO9tZPxuoMD+dVnTJElkkEnYUADikNdmhvsceU6GXXsVh8VOeYm1tjOtAWRT7KrpN0sWS9rR3tvcP/5rEbAAkZdNp1icAUHXiyaQ0MKDeBx8KO8qU5PJ5+ek0i5ahbIodU/DqSU0BYES5Z59Vbs8eBhkDqDrTzjpLNm2aeru6dMwrXxF2nCmnf8sWOd9XlDEdKJOiioLuNd2/mewgAI6UHRpk3LYs5CQA8EINkYjiK1eqZy3jCsLge54kBhmjfIptKVB7Z/tbJa2RdHz3mu7T2zvbXyZpdvea7v+ctHTAFOd7g0UBLQUAqlE8mVTPb36j/m3bFFm4MOw4U4q/PiWLRpmEApIkry1x6H16Iu2d7rUlXiZpdiLtFf0+vagxBe2d7R+R9BlJP5c09FO/R9IVpUUGUAo/k1bT8ceracaMsKMAwBFaVyclST3MQlRxfiql6LJlsqaiP99FnfLaEmV5n17sQOP3SDq/e033lyW5YN8GSaeUcjEApcmmM4rSdQhAlWpetEjN8+ert2tt2FGmFOecfM9j0TIMeY+k8xNpb0Lv04stCmZ2r+neEDweupgNewygzAr9/cpu3qzYMroOAahOZqZ4skN9Dzwg18+EhJUysGOHCgcOMJ4AQ2Ym0t6E36cXWxSk2jvb//qwfedJeryUiwEoXv+mTVIux0rGAKpaazKpQl+f+tatCzvKlOGngkHGCYoCSJJSXltiwu/Ti+2IdpWk/27vbP+BpGh7Z/vXJb1R0uEBAJSJn85IkqK0FACoYi2rVklNTertWqv4ypVhx5kS/FRKampS9NSlYUdBdbhK0n97bYkfSIp6bYlxvU8vqqWge033vZJWSToo6dfB9728e033gyVFBlC0bCYti8UUWcSMHgCqV2Nrq6atOEO9DDauGD+VUvSUU9QQjYYdBVUgkfZGfJ+eSHslvU8vqqWgvbP9L7vXdN8l6f2H7b+ie033F0q5IIDi+F5a0VNPlTU2hh0FAEbVmkxqz1e+qtzevWqaNSvsOHXNOSc/lVLry14WdhRUAa8t0STpJ5Jel0h77x/r+NEUO6bgxvbO9hXDd7R3tr9bhxUJAMrDOSc/k1GM9QkA1IB4crUkqXctsxBNttzuPcrv3csgY0iSEmkvJ+klknITPVexRcE7Jf2kvbN9vnRoIbNPS3rNRAMAOFLuqadU2L+f6UgB1ITY8oQaZ85kvYIK8FPrJUmx0ygKcMitki6f6EmK6j7Uvab77vbO9s9L+nl7Z/sXJX1R0l90r+lOTzQAgCMNrWRMSwGAWmANDYqfe656194nVyjIGor9zBGl8j1PMlNsGR8a4ZCzJH3Qa0u8T9JWSYWhJxJp7y+KPUnRy+B1r+n+dntn+xJJ/6bBguCxEsICKEE2M1gURE/llz6A2hBPduj5n/1M2XSari2TyE+lFFm8WA3xeNhRUD1+G3xNyFGLgvbO9o06ctGDhmBfZ3tnuySpe033qRMNAeCF/HRGzQsXqrGVX/oAakNrR4ckqadrLUXBJPJTKbWsODPsGKgiibT3mXKcZ7SWgn8uxwUAlC6bTtM0DKCmNM2Zo2hbm3q7ujT7skvDjlOXcvv2KbfrScXe8pawo6DKeG2JBZLeLGmBpO2SbkukvR2lnOOoRUH3mu7OicUDMB6F3l71b9um6RdeEHYUAChJa7JDeztvUb6nl5bOSZD1gpWMaYnBMF5bIinpbkm/l/RHSWdK+oTXljg/WMOgKEWPKWjvbJ8l6aWS5kiyof3da7pvKfYcAMbmb9ggOccgYwA1J55crb033qS+hx7UMa98Zdhx6o4/VBQkEiEnQZX5gqQPJNLezUM7vLbEOyRdq8FFzYpS1PQA7Z3tr5a0WYNTHt0s6UuSbpL0TyUEBlCEbCYjSXQfAlBzWs46U9bSwurGk8Rfn1LzSSep8bjjwo6CCTCz88wsY2abzOzKEZ6/yMz2mNm64OuSMU6ZkPTdw/bdIqmkNxLFzhl2taTPdq/pniOpJ/jzc5KuL+ViAMbme2k1TJ+uppNOCjsKAJTEIhHFV65UTxeLmE0GP5VifYIaZ2aNkr4h6XxJyyW9ycxG+kf9vnNuRfB14xinfVqD05IOd5ak3aVkK7YoWCrpK8Hjoa5D10j6UCkXG4mZHWdmPzSztJl5ZvZnZjbTzP7HzDYGf84IjjUz+1pQWf3ezA6/AUDNGxpkbGZjHwwAVSaeTGpg2zb1b9sWdpS6ku/pVf/WrYrSdajWrZS0yTm32TnXL+kOSRdO8JxflXSX15b4nNeWeKfXlvispJ8F+4tWbFHQJykaPN7b3tm+UFJE0oxSLnYUX5V0t3OuTdIZkjxJV0r6pXNuqaRfBtvSYFW1NPi6TNK3ynB9oGq4QkH+xo2KMp4AQI1qTQ5NTUoXonLKZtKD480YZFzr5mlwdqAhO4J9h3td8AH4D81swWgnTKS9b0n6oAYLjr+XdI6kDyXS3jdLCVbsQOP7JL1W0u2Sfi7pTklZSfeXcrHDmdmxkl4m6SJJCiqmfjO7UNLLg8M6Jd0j6WMarKRucc45SQ8ErQwnOueenEgOoFoMbNsm19enWBvjCQDUpuZFi9Q8f756u9Zq5pvfHHacuuGvT0li5qFqN6OxsclrSzwybNcNibR3Q4mn+amk/3DOZc3sXRp8LzzqyP1E2vsPSf9R4nVeoNii4K36U6vC30v6qKRjJH15IheXdLKkPZK+Y2ZnSHpUg5XO8cPe6D8l6fjg8dGqK4oC1AU/HaxkTEsBgBplZoOrG9/5U7n+flkkEnakuuB7nhpnz1bz3LlhR8Eo9uXzuUTaO3uUQ3ZqcC2BIfODfYc45/YO27xRg7MLHZXXlnivpAcTae/RYfvOlnR2Iu0VPf632O5DL+1e090rSd1ruv3uNd3/0r2m+0qVOKp5BE0aHAjxLefcmZJ69aeuQpKkoFXg8JWVR2Vml5nZI2b2SC6Xm2BEoHL8dFpqbFT0lFPCjgIA49aaTKrQ16e+x9aFHaVu+KmUYssZT1AHHpa01MxONrOIpDdqsAfOIWZ24rDNCzTYtX40H5O05bB9W3TYe+qxFFsU/Owo+39SysVGsEPSDufcg8H2DzVYJDw9dEOCP4dGT49ZXUmSc+4G59zZzrmzm5qKXooBCF02nVF0yclqiEbHPhgAqlTLqlVSUxNTk5ZJIZtVdtMmxRJ0Hap1zrmcpMsl/UKDb/Z/4Jxbb2afNbOhVUs/YGbrzexxSR9Q0M1+FMcl0t6zh+17VtKsUrIVWxQcMQ1Ke2f7MZIKpVzscM65pyRtN7OhFodXSUppsGJaE+xboz8VH3dKenswC9EqSfsZT4B64qfTii6j6xCA2tbY2qqWFSvUs5aioByyGzZK+TzjCeqEc+4u59ypzrkXOef+Jdj3SefcncHjf3TOneacO8M59wrnXHqMU2712hJ/dti+cySVNAXYqB+jt3e2b9Rg151p7Z3tGw57eq6k/ynlYkfxfkm3BU0omyW9Q4PFyg/M7GJJWyW9ITj2Lkl/KWmTBmdEekcZrg9Uhfxzzyn31FOKJSgKANS+eDKpPV/5inLPPKOm2bPDjlPT/FQwyJg1CjCyf5P0/7y2xD9L2qjBWTqvkvSvpZxkrL41/6zBVoJvSfqXYfsLGhwA/OtSLjYS59w6SSMNyHjVCMc6Se+b6DWBauSnB1cypqUAQD0YKgp677tPx15wwdjfgKPyvZQapk9X87yRZq7EVJdIezd4bYlGDX7QvljSE5KuDqYqLdqoRUH3mu5OSWrvbE93r+l+YHxRARQjmxlsHWQ6UgD1ILY8ocaZM9XT1UVRMEF+ylMskWBRSxxVUABMaP2uokbhUhAAk8/30mqcM5tmdgB1wRoaFD/3XPWuvU+uUJA1FDuMEcO5XE7ZTEYzWPMBRfDaEi+XtEJSVyLtPTLG4S/ATyhQJfxMRjG6DgGoI/Fkh/J798r3xppREUeT3bxZLptlOlIcwWtL/LvXlrh02Palkn4l6Z8k3e+1JV5byvkoCoAq4Pr71b9pE12HANSV1o4OSVJv19qQk9SuQ4OMmXkIR3qFpP8dtv1Pkj6WSHuzNTiV6d+XcjKKAqAKZLdskRsYYJAxgLrSNGeOookE6xVMgJ9KyaZNU2Tx4rCjoPrMSaS9LZLktSWWSDpJ0tAKxjerxEWGiy4K2jvbj23vbH9ze2f7FcH2Ce2d7SeVcjEAI8umg0HGTEcKoM60JjvU99hjyvf0hh2lJmVTnmLLlskaG8OOguqT9doSQ6udni0pk0h7B4LtAUklrYRaVFHQ3tn+Eg2uDXClpE8Eu0+X9PVSLgZgZH46I4tE+CQIQN2JdySlXE59Dz0YdpSa4woF+Z5H1yEczSOS3u+1JWKS3q4Xrh+2RNKeUk5WbEvBVyRd0b2m+3RJuWDffZJWlXIxACPLZtKKLl0qaypqQjAAqBktZ50pa2mhC9E4DGzbpkJvL4OMcTRXSfqYpF5JZ0n68rDn3iDp/lJOVmxRcJqk7waPnSR1r+nukRQv5WIAjuSck++lFaXrEIA6ZJGI4itXqofBxiUbmrWJlgKMJJH21mlwsbKXSjo1kfa2D3v6R5KuKOV8xX4suUfSQklbh3a0d7afImlnKRcDcKTc7j3K79vHdKQA6lY8mVTPPfeof+tWRRYtCjtOzfBTKam5WdFTTgk7CqpUIu31SvrdCPszpZ6r2JaCTkl3tHe2JyVZMMbgRkn/XuoFAbwQKxkDqHetq5OSpB66EJXET3mKLj1FFomEHQVTQLFFwTUaXAzhLknHSvq1pHslfW2ScgFThu8NFgXRZRQFAOpTZNEiNS9YwHoFJXDOyU+l6DqEiimq+1D3mu68pI9L+nh7Z/vs7jXdz0xuLGDqyGbSap43T43Tp4cdBQAmTTzZoed/cqdcfz+ffBch99RTg11LKQpQISUvXkZBAJSXn84o2sZ4AgD1rTWZVKGvT32PrQs7Sk04NMg4wcxDqIyiWgraO9uXabCr0NmSjhn+XPeabsp9YJwKBw+q/4knNP2888KOAgCTquWcc6SmJvV2dSl+zsqw41Q9f31KamhQjK6lGAevLdEo6eOJtPfZYr+n2JaC70naLemtkl5z2BeAccpu3CgVCkxHCqDuNba2qmXFCvWsZbBxMfxUSpElJ6uhpSXsKKhNTZI+Veo3FGOZpFXB2AIAZeKnh2YeoigAUP/iyaT2fOUryj3zjJpmzw47TlXzPU8tL31p2DFQxby2xJtHebq51PMVWxQ8LOlFkjaUegEAR5dNZ9QQj6t53rywowDApIuvHiwKeteu1bEXXhh2nKqV27tXuaeeYpAxxvI9SdslFUZ4zko9WbFFwTsk3dje2f4LSU8Of6J7TfftpV4UwCA/nVZ02TJZQ8lj/gGg5sQSCTXOnKmeLoqC0fgpBhmjKNskvSmR9u4//AmvLRGT1FvKyYotCl4n6ZWSzpDUN2y/k0RRAIyDKxSUzWT4jxHAlGENDYp3dKh37Vq5QoEPRI7i0MxDyykKMKp1klZIOqIo0OB79JJaC4otCq6S9Nfda7rvLuXkAI5uYOdOFXp7FWUlYwBTSGuyQ8//9KfyPU/TTjst7DhVyU+l1LxgAevXYCzv1chdh5RIe1mVuPRAsUWBk/SLUk4MYHSHPglikDGAKSTe0SFJ6u1aS1FwFL6XousQxpRIe7vKeb5iK4ibJV1UzgsDU102nZEaGhRdujTsKABQMU2zZyuaSKi3i6lJR5I/cEADW7cxyBhj8toSNxy2PaEFQIptKThb0ofbO9s/oiMHGv/FRAIAU5WfySiyeLEapk0LOwoAVFRrskN7v/Nd5Xt61NjaGnacqnKoFfk0igKM6Y2SLhu2fbekmeM9WbFFwb3BF4AyyabTmnbG6WHHAICKiydXa++/36i+Bx/UMa96VdhxqkrWY+YhFO3wgcQlT0M6XFFFQfea7s9M5CIAXij//PMa2LlTx73hDWFHAYCKazlzhRpaWtTT1UVRcBg/lVLT3Lks7oZiuDG2S3LUoqC9s/2E7jXdTwWPTzracd1russ6yAGYCrKZjCQplmCQMYCpxyIRtZxzjnq71oYdper4qRTjCVCsiNeWuGrYduywbSXS3ueLPdloLQUbJA3NhbVDR1YfFuxrLPZiAAb56cGiILqMogDA1BRPdqjn179W/9atiixaFHacqlA4eFDZP27WMa95TdhRUBsekDT8xfLgYdtOUlmKguHzhJ1c7AkBjM1Pe2qcMUNNc+eEHQUAQtGaTOppST1dXZpJUSBJym7YIBUKtBSgKIm09/Jynu+oRUH3mu7t7Z3tz3ev6Z7evaZ7azkvCkx12XRGsUSbzCY0JggAalZk0SI1L1ig3q61mvmWt4Qdpyr4qZQkURQgFGOtU8A7FqDMXC6n7MaNdB0CMOXFkx3qffBBuf7+sKNUBT/lqfHYY9V04olhR8EUNFZRMKFRzACO1P/EE3L9/Yq1LQs7CgCEqnX1arm+PvX97rGwo1QFP5VS7LTltCIjFGNNSRpr72y/ebQDutd0v7OMeYC653tpSVK0jTmoAUxtLSvPkZqa1Lu2S/FV54QdJ1RuYEDZDRs04+1vCzsKpqixWgokKT/GF4ASZDNpWXOzoksYvw9gamtsjavlzDPVw9Skyv7xj3IDA4wnQGjGainwu9d0X1qRJMAU4aczipxyiqy5OewoABC6eDKpPdddp9wzz0zpBbv89QwyRriKaSkAUEZ+Oq3YMsYTAIA0ONhYknrXTu3WAt/z1NDSwpoNCA2zDwEVlHvmGeWfeYaVjAEgEEsk1Dhz5pTvQuSnUoomErIGPq9FOEZ95XWv6T6mUkGAqYCVjAHghayhQfGODvV2dckVCmHHCYXL5wdbkek6hBBRjgIVlM0MzjzEdKQA8CetyQ7l9+2Tn/LCjhKK/q3b5Pr6FEswKx3CQ1EAVJDvpdV04olqPO64sKMAQNWIdwTjCrq6Qk4SjkMrGZ9GSwHCQ1EAVFA2wyBjADhc0+zZii5PTOmiwCIRRZcsCTsKpjCKAqBCCtmsspu3KErXIQA4QmtHUn3r1inf0xN2lIrzvZSip57KVNUIFUUBUCHZjZukfF6xNgYZA8Dh4smklMup78EHw45SUc45+SmPQcYIHUUBUCF/GmRMUQAAh2s5c4UaWlrUM8W6EOV27VJh/37FljPIGOGiKAAqxE9nZC0tal64MOwoAFB1LBJRyznnqPfeLjnnwo5TMQdTrGSM6kBRAFRINp1WbOlSFqYBgKOIJzs0sGOHBrZuDTtKxfiplNTYqOipp4YdBVMc706ACnDOyU+nFWUlYwA4qtbVqyVpSq1unE15ii5ZooZYLOwomOIoCoAKyO3apcKBA4wnAIBRRBYuVPPChVNqalI/laLrEKoCRQFQAX4mI0mKskYBAIyqNdmh3ocekuvvDzvKpMvt2aPcnj0sWoaqQFEAVIDveZKZYvQZBYBRxZNJub4+9f3usbCjTDrf8yRJsQQzDyF8FAVABWTTGUUWLlRDPB52FACoai0rz5GamtS7tv67EPnBzENRigJUAYoCoAL8TEZRxhMAwJgaW+NqOfNM9dw7BYqC9SlFFi1SY2tr2FEAigJgsuV7ejWwbZtibYwnAIBixJNJZdNp5fbsCTvKpPI9T1EWLUOVoCgAJll2QzDImJYCAChK6+qkJKlnbf1OTZrfv18DO3Yw8xCqBkUBMMn8dFqSmI4UAIoUbWtT46xZ6q3j9Qp8L/i/IUFRgOpAUQBMsmw6o4Zjj1XTCSeEHQUAaoI1NCjeca56166VKxTCjjMphgYZx+g+hCpBUQBMMj+dVmzZMplZ2FEAoGa0JpPK79snP+WFHWVS+KmUmk48UU0zZ4YdBRVmZueZWcbMNpnZlaMc9zozc2Z2diVyURQAk8jl88pu2KBYgq5DAFCKeEeHJNXt6sa+57E+wRRkZo2SviHpfEnLJb3JzI7oQ2Zmx0j6oKQHK5WNogCYRP1bt8n5vqLLKAoAoBRNs2Ypujyhnq57w45SdoW+PvVv3swg46lppaRNzrnNzrl+SXdIunCE4z4n6RpJfqWCURQAkyibGRpkzHSkAFCq1o6kDq57XPmenrCjlJWfzkjOURRMTfMkbR+2vSPYd4iZnSVpgXPuvysZjKIAmES+l5aamhQ55ZSwowBAzYmvTkq5nPoeeCDsKGXlewwyrlczGhubvLbEI8O+Livl+82sQdKXJX10chIeXVOlLwhMJX4mreiSJWqIRMKOAgA1p2XFCjW0tKinq0vHvPrVYccpGz+VUuPMmWo6/viwo6DM9uXzuUTaG21g8E5JC4Ztzw/2DTlG0osl3RNMUHKCpDvN7ALn3CPlzjscLQXAJMqmM4rSdQgAxsUiEbWsWqXerrVyzoUdp2z8lKfY8uXMSjc1PSxpqZmdbGYRSW+UdOfQk865/c652c65xc65xZIekDTpBYFEUQBMmty+fco9/bRiDDIGgHGLJzs0sGOHBrZuDTtKWRT6+5XdtImZh6Yo51xO0uWSfiHJk/QD59x6M/usmV0QZja6DwGTJDu0kjHTkQLAuLUmk3paUk/XWs1cvDjsOBOW3bhRGhhQ7DQGGU9Vzrm7JN112L5PHuXYl1cik0RLATBp/HRGkhRtoygAgPGKLFyo5oUL1XtvfUxNmvUGF2OjpQDVhqIAmCTZtKemOXNYrRIAJqg12aHehx5Sob8/7CgT5qdSamhtVfOCBWMfDFQQRQEwSfx0RlG6DgHAhMWTq+UOHtTB3/0u7CgT5q9PKZZIyBp4C4bqwisSmASuv1/ZzZsZZAwAZRA/Z6XU3Kzerq6wo0yIy+flZzKsT4CqRFEATILs5s3SwADTkQJAGTTE42o580z1dK0NO8qE9G/ZIuf7rGSMqkRRAEwC3xuaeYhPgwCgHOLJpLLptHJ79oQdZdz81NBKxhQFqD4UBcAkyKbTslhMkUWLwo4CAHWhNdkhSepZW7utBX7Kk0Wjipx8cthRgCNQFACTwM9kFF26VNbYGHYUAKgL0bY2Nc6apd57a3dcgZ9KKdq2TNbEMlGoPhQFQJk555T1PMUYTwAAZWMNDYp3nKve++6TKxTCjlMy55x8z6PrEKoWRQFQZrmnn1Z+/34WLQOAMmtdvVr5ffvkr0+FHaVkAzt2qHDgAGPNULUoCoAy89PBIGOKAgAoq/i550qSetfWXheioUImtvy0kJMAI6MoAMosm85IkqLL6D4EAOXUNGuWYsuXq6cG1yvwPU9qalL01KVhRwFGRFEAlJmfTqt5wQI1traGHQUA6k48mdTBdY8r39MTdpSS+KmUoqecooZIJOwowIgoCoAyy6bTDDIGgEkST3ZIuZz6Hngg7ChFc87JT6UYZIyqRlEAlFGhr0/9W7cquozxBAAwGVpWrFBDS4t6amhq0tzuPcrv3csgY1Q1igKgjLIbNkjO0VIAAJPEIhG1rFql3q4uOefCjlMUP7VekhQ7jZYCVC+KAqCM/KFBxm18GgQAkyWe7NDAzp3qf+KJsKMUxU+lJDPFmIACVYyiACgjP5NWwzHHqHneSWFHAYC61bp6tSSpt2ttyEmK43ueIosXqyEeDzsKcFRVURSYWaOZPWZmPwu2TzazB81sk5l938wiwf5osL0peH5xmLmBw2XTGUWXnSozCzsKANStyIIFal60UL01MjUpg4xRC6qiKJD0QUnesO1rJF3nnDtF0j5JFwf7L5a0L9h/XXAcUBVcoSA/k1GMrkMAMOlaO5LqfeghFfr7w44yqty+fcrtepKiAFUv9KLAzOZL+itJNwbbJumVkn4YHNIp6bXB4wuDbQXPv8r4SBZVYmD7drm+PgYZA0AFxJNJuYMHdfB3vws7yqiy3uBnnrHlfGCE6hZ6USDpK5KukFQItmdJes45lwu2d0iaFzyeJ2m7JAXP7w+OB0J3aJAx05ECwKSLn7NSam5Wz733hh1lVH4qJUlMR4qqF2pRYGZ/LWm3c+7RMp/3MjN7xMweyeVyY38DUAZ+2pMaGhRdekrYUQCg7jXE42o588yqH2zspzw1z5unxuOOCzsKMKqwWwo6JF1gZk9IukOD3Ya+Kuk4M2sKjpkvaWfweKekBZIUPH+spL2Hn9Q5d4Nz7mzn3NlNTU2HPw1Mimw6o8iSk9UQi4UdBQCmhHgyqWwmo4Hdu8OOclSDg4xpJUD1C7UocM79o3NuvnNusaQ3SvqVc+4tkn4t6fXBYWsk/SR4fGewreD5X7laWbkEdc/PpBWj6xAAVEzr6qQkqXftfSEnGVm+p1f9TzzBIGPUhLBbCo7mY5I+YmabNDhm4KZg/02SZgX7PyLpypDyAS+Q379fuV1PKsogYwComOiyZWqcPbtqpybNZtKSpCjjCVADqqZvjXPuHkn3BI83S1o5wjG+pP9b0WBAEYYGGTMdKQBUjjU0qLXjXPX89l65QkHWUF2fdfrrg0FNrelAAAAgAElEQVTGtBSgBlTXTw9Qo4Y+DWI6UgCorHgyqfy+fYfegFcTP5VS45zZap47N+wowJgoCoAy8NMZNc6apaY5c8KOAgBTSvzccyVJvV3VNzWp73lMRYqaQVEAlIGf9hRbRisBAFRa06xZii1frp4qm5q0kM0qu2kTXYdQMygKgAlyAwPq37hJ0QQzDwFAGOLJpA6uW6f8gQNhRzkku2GDlM9TFKBmUBQAE5TdskVuYECxNooCAAhD6+qklM+r94EHwo5yiJ/yJDHIGLWDogCYoGxmcOahKN2HACAU01asUEM8XlWrG/uplBqmT1fzvHlhRwGKQlEATJDvpWWRiKInnxx2FACYkqy5WS2rVqm3q0vVsqap73mKLV8uMws7ClAUigJggrLptKKnnCJrbg47CgBMWa3JDg3s3Kn+J54IO4rcwICy6TQzD6GmUBQAE+Cck59OK8p4AgAIVTyZlCT13hv+6sbZzVvk+vsZT4CaQlEATEBuzx7ln32WRcsAIGSRBQvUvGihetaGXxT43tBKxrQUoHZQFAATcGiQMS0FABC61o6k+h56WIX+/lBz+KmUbNo0RRYvDjUHUAqKAmAC/HRakli4DACqQHx1Uu7gQR189NFQc/iplGJtbbLGxlBzAKWgKAAmIJvOqOmkE9V47LFhRwGAKS++cqXU3KyervC6ELlCQVmPQcaoPRQFwAT46bRibfziB4Bq0BCPq+Wss0Jdr2Bg2zYVensVO41BxqgtFAXAOBV8X/1btjDIGACqSDzZoWwmo4Gnd4dyfT81NMiYogC1haIAGKfsxk1SoaDoMgYZA0C1aB2amnRtOK0FvudJzc2KvuhFoVwfGC+KAmCc/LQnSbQUAEAViS5bpsbZs9Ub0rgCf31KsaVLZZFIKNcHxouiABinbDqjhpYWNS9YEHYUAEDAGhrU2nGueu+7Ty6fr+i1nXPyPY/xBKhJFAXAOPmZtKLLlska+DECgGoST65W/rnnDvXvr5TcU08pv2+fosw8hBrEuxlgHJxzyqYzitJ1CACqTrzjXMms4l2IhoqQaQwyRg2iKADGYWDnThV6epiOFACqUNPMmYotX66eCk9N6qc8qaFBURa0RA2iKADGITu0kjEtBQBQleLJpA6uW6f8gQMVu6afSimy5GQ1TJtWsWsC5UJRAIyDn85IZoouXRp2FADACFqTHVI+r97776/YNf1UivUJULMoCoBx8NOeIosXq6GlJewoAIARTFuxQg3xeMVWN87t3avc008rlqAoQG2iKADGgUHGAFDdrLlZLatWqberS865Sb+enwrWrqGlADWKogAoUf7AAQ3s2KEYKxkDQFVrXZ3UwK5d6t/yxKRfa2jmoViC/xtQmygKgBJlMxlJoqUAAKpcPJmUpIpMTep7npoXLFDj9OmTfi1gMlAUACXy04NFQYzFaQCgqkXmz1dk0SL1rK1AUcAgY9Q4igKgRNlMWo3HHaemuXPDjgIAGEM8mVTfQw+rkM1O2jXyBw5oYNs2igLUNIoCoER+OqNoW5vMLOwoAIAxxJMdcgcP6uCjj07aNXxvaJAxLcioXRQFQAlcLqfshg2KtTGQDABqQXzlSqm5eVJXN/7TIGOKAtQuigKgBP1bt8plswwyBoAa0RCPq+WssyZ1sHHW89Q0d66aZs+etGugfpjZeWaWMbNNZnblCM+/28y6zWydmXWZWUX6pVEUACXw02lJoqUAAGpIPNmh7IYNGnh696Scn0HGKJaZNUr6hqTzJS2X9KYR3vTf7pxrd86tkPQFSV+uRLamSlwEqBfZdFpqblZ0yZKwoxzhHd99f9gRQvGdi74edgQAVa519Wrt+dKX1bt2rY77P39b1nMXDh5U9o+bdcxr/qKs50XdWilpk3NusySZ2R2SLpSUGjrAOff8sOPjkiZ/9T3RUgCUxE9nFH3Ri2SRSNhRAABFii5bpsY5syelC1F2wwapUGCQMYo1T9L2Yds7gn0vYGbvM7M/arCl4AOVCEZLAVCCbDqt+Lnnhh0DAFACM1PruR3q+c1v5PJ5WWNj2c59aJAx3YcgaUZjY5PXlnhk2K4bEmnvhlLP45z7hqRvmNmbJf2TpDXlyng0FAVAkXJ79yq3Z4+ijCcAgJoTTya1/yc/kb9+vaadfnrZzuunUoNr15x4YtnOidq1L5/PJdLe2aMcslPSgmHb84N9R3OHpG+VI9tY6D4EFOnQIOMERQEA1Jp4x7mSmXrK3IXIT3mKLU+wdg2K9bCkpWZ2splFJL1R0p3DDzCzpcM2/0rSxkoEoygAipRNZyQN9k0FANSWppkzFVu+XL1lXK/A9fcPrl1D1yEUyTmXk3S5pF9I8iT9wDm33sw+a2YXBIddbmbrzWydpI+oAl2HJLoPAUXzM2k1HX+8mmbMCDsKAGAc4smk9t54o/IHDqjxmGMmfL7sH/8oNzBAUYCSOOfuknTXYfs+OezxByseSrQUAEXLemkWLQOAGta6Oinl8+q9//6ynM9PeZKkKCsZow5QFABFKPT3K7tli2Jt/OIHgFo17Ywz1BCPl60LkZ9KqaGlRZFFi8pyPiBMFAVAEfo3bZJyOcVoKQCAmmXNzWr5s1Xq7eqScxNfD8r3PEWXJ2QNvJ1C7eNVDBTBPzTImJmHAKCWtSaTGti1S/1btkzoPC6fl59OK5ZgPAHqA0UBUAQ/7cmmTVNk0cKwowAAJiCeTErShFc37t+6Va6vj0HGqBsUBUARsumMoqcuLesqmACAyovMn6/IokUTXq9gaJBxbDljzVAfKAqAMTjn5GcyitF1CADqQjyZVN9DD6uQzY77HH4qJYtEFF2ypIzJgPBQFABjyD35pAr79zMdKQDUifjqpJzv6+Cjj477HH4qpeiyZbLm5jImA8JDUQCMYWiQMdORAkB9iK9cKWtuVs84pyZ1zsn3PMVYnwB1hKIAGEM2k5YkRU89NeQkAIByaGhp0bSXvGTcg40Hdu5SYf9+BhmjrlAUAGPw0xk1L1yoxtZ42FEAAGXSmuxQdsMGDTz9dMnf66fWS5Jip1EUoH5QFABj8NOeYm0MMgaAevKnqUlL70Lke57U2EgLMuoKRQEwikJvrwa2bWeQMQDUmeiyZWqcM1u9a0vvQuSnUoq+6EVqiEYnIRkQDooCYBT+hg2Sc7QUAECdMTO1ntuh3rX3yeXzJX1vNuUxngB1h6IAGEU2PTjIOLaMlgIAqDfx1auV379f/vr1RX/PwO7dyu3Zw6JlqDsUBcAo/HRGDdOnq+mkk8KOAgAos3jHuZJZSasbZ72hlYxpKUB9oSgARpFNpxVbtkxmFnYUAECZNc2Yodhpp5U02NgPioIo3UpRZygKgKNw+bz8jRv5xQ8AdSye7NDBxx9X/vnnizreX59SZNEiNba2TnIyoLIoCoCj6N+2Ta6vj0HGAFDHWpNJKZ9X7/0PFHW8n0qxPgHqEkUBcBTZTEaSmI4UAOrYtDPOUEM8XtTqxvn9+zWwc6eiCQYZo/5QFABH4afTg4vTnHJK2FEAAJPEmpvV8mer1LO2S865UY/1GWSMOkZRABxF1ksruuRkFqcBgDrXmlyt3K4n1b9ly6jH+etTkigKUJ8oCoCj8DMZRdtoIgaAehdPJiVpzC5Evuep6cQT1TRjRiViARVFUQCMIP/cc8o99ZRijCcAgLoXmT9PkcWLx1yvwE+laCVA3aIoAEbgp4NBxsuYeQgApoJ4Mqm+hx5WIZsd8flCX5/6t2xhJWPULYoCYAR+OhhMRksBAEwJ8WSHnO+r75FHRnzeT2ck5xRL0FKA+tQUdgCgGmXTGTXOma2m2bPDjgIAU96v//CxSb+GteY0t6lB6f/6Vx049rQjnm/55RZNl/S7xh+r8Ie7Jz2PJL3ixddU5DqAREsBMCI/k1GMrkMAMGW4WJP622Yq8vjuEZ9v2rJf+ekRFWbGKpwMqAyKAuAwrr9f2U2bFEtQFADAVNK/Yo6atx1Qw96DRzzXvPk55ZYcK5mFkAyYfBQFwGGyW7ZIAwMMMgaAKSa7Yq4kKfr4nhc+MZBX0/YDGjj5uBBSAZVBUQAcJptOS2KQMQBMNblF05WfEVVk3Qu7EDVtPyDLOw2cfGxIyYDJR1EAHMZPZ2SRiCKLF4cdBQBQSWbqP2Ouor/fI+Xdod3Nm/dL0mD3IaBOURQAh/HTnqKnniprYnIuAJhqsivmqKFnQM1/fO7QvuYt+1VoaVL++JYQkwGTi6IAGMY5p2w6oyhdhwBgSsqePkfO9IJZiJq27FduMYOMUd8oCoBhcrv3KL9vH9ORAsAU5aZHlVtynKJD4wryTs1PPK8Bug6hzlEUAMNkh1YyZjpSAJiysivmqHnjc7LeATXu6pH15ykKUPcoCoBh/HRGkhRdRvchAJiqsivmygpOke49at4yOLYgx8xDqHOMpASGyWbSap43T43HHBN2FABASAaWzlChpUnRdXvkYo1ykUblTmoNOxYwqSgKgGF8L61oG12HAGBKa2pQ/4tnK7put/JzWzSwaLrUSOcK1Dde4UCgcPCg+rduVYyiAACmvOyKuWp85qCaM8+yPgGmBIoCIJDduFEqFJiOFACg/hVzJYmVjDFlUBQAAT+dliRaCgAAys9tUe6kuCRRFGBKoCgAAtl0Wg2trWqeNy/sKACAKpB9yfEqRBuVW8jkE6h/oRYFZrbAzH5tZikzW29mHwz2zzSz/zGzjcGfM4L9ZmZfM7NNZvZ7MzsrzPyoH70PPKD9//UTTTvjDFkDtTIAQOp5Q5v2fuFlUnNj2FGASRf2u5+cpI8655ZLWiXpfWa2XNKVkn7pnFsq6ZfBtiSdL2lp8HWZpG9VPjLqzfN3/0LbL71MzfPm6cTPfz7sOACAKuGmNSk/j1YCTA2hFgXOuSedc78LHh+Q5EmaJ+lCSZ3BYZ2SXhs8vlDSLW7QA5KOM7MTKxwbdWTf93+gnR/+sGIvfrEWfe9WNR8/N+xIAAAAFRd2S8EhZrZY0pmSHpR0vHPuyeCppyQdHzyeJ2n7sG/bEewDSuKc0zPXf1tPfepTir9stRbefJMaj2UgGQAAmJqqYvEyM2uV9CNJH3LOPW9mh55zzjkzcyWe7zINdi9SJBIpZ1TUAVcoaPc11+jZzls0/YK/0Un/8i+y5uawYwEAAIQm9JYCM2vWYEFwm3PuP4PdTw91Cwr+3B3s3ylpwbBvnx/sewHn3A3OubOdc2c3NVVF3YMq4QYGtOvKK/Vs5y2a8fa36aSrr6YgAAAAU17Ysw+ZpJskec65Lw976k5Ja4LHayT9ZNj+twezEK2StH9YNyNgVIWDB7Xj8vfr+Tt/qjkf+qCO/8d/ZKYhAAAAhd99qEPS2yR1m9m6YN9Vkq6W9AMzu1jSVklvCJ67S9JfStokqU/SOyobF7Uq//zz2v7u9+jgY4/phE9/WjPe+HdhRwIAAKgaoRYFzrkuSXaUp181wvFO0vsmNRTqzsDu3dp+yaXKbtmiedd9WdPPOy/sSAAAAFUl7JYCYFL1b92qbRdfotyzz2rht69X/Nxzw44EAABQdSgKULd8z9O2Sy+Tcjkt6vyuprW3hx0JAACgKjHKEnWp7+GHtfVtb5c1N2vR7bdREAAAAIyCogB158CvfqVtl1yqprlztfj22xRdsiTsSAAAAFWNogB15bn//LF2vP8Dii5bpkW3fU/NJ54YdiQAAICqR1GAurH3ppv15FVXKX7OOVr0nZvVNGNG2JEAAABqAkUBap5zTru/+EXtvvZaHXP+eZp//bfUEI+HHQsAAOAIZnaemWXMbJOZXTnC8x8xs5SZ/d7MfmlmiyqRi6IANc3lcnryE5/Q3htv0nFv/DvN++IX1RCJhB0LAADgCGbWKOkbks6XtFzSm8xs+WGHPSbpbOfc6ZJ+KOkLlchGUYCaVchmteNDH9L+H/5Is9/7Xp3wqU/JGhvDjgUAAHA0KyVtcs5tds71S7pD0oXDD3DO/do51xdsPiBpfiWCsU4BalK+p0c73vs+9T30kI7/+Mc1821vDTsSAADAWOZJ2j5se4ekc0Y5/mJJP5/URAGKAtSc3N692nbppcpu2KiTrr1Wx/7NX4cdCQAAQDMaG5u8tsQjw3bdkEh7N4znXGb2VklnS/rzsoQbA0UBakr/jp3advE7lXt6txZ88xtqfdnLwo4EAAAgSdqXz+cSae/sUQ7ZKWnBsO35wb4XMLNXS/q4pD93zmXLm3JkFAWoGf6GDdp+8SUq9Pdr4XduVsuZZ4YdCQAAoBQPS1pqZidrsBh4o6Q3Dz/AzM6U9G1J5znndlcqGAONURP6fveYtr71bZKZFt16CwUBAACoOc65nKTLJf1CkifpB8659Wb2WTO7IDjsWkmtkv6fma0zszsrkY2WAlS9nt/+Vjs+8EE1H3+8Ftx0kyLz54UdCQAAYFycc3dJuuuwfZ8c9vjVFQ8lWgpQ5fb/9Kfa/t73KbLkZC26/TYKAgAAgElAUYCq9ewtt2rXP1yhlrPO0qJbblHTrFlhRwIAAKhLdB9C1XHO6Zmvf13PfPNban31qzTvS19SQzQadiwAAIC6RVGAquLyeT31uc/puTu+r2Nf/zqd+OlPy5p4mQIAAEwm3m2hahT6+7Xrio/pwN13a9all2jORz4iMws7FgAAQN2jKEBVKPT2asf736/e++7X3Cuu0Kx3viPsSAAAAFMGRQFCl9u3T9sve5f8VEonfv7zOu7//G3YkQAAAKYUigKEauDJJ7Xt4ks0sHOn5n/9azrmla8MOxIAAMCUQ1GA0GQ3b9a2iy9R4cABLbzx39Xy0peGHQkAAGBKoihAKA7+/vfaftm7pKYmLbr1FsUSibAjAQAATFksXoaK673vPm296B1qaG3V4ttvoyAAAAAIGUUBKur5u+/Wtne9W5H587Xo9tsUWbgw7EgAAABTHkUBKmbfHXdo54c/omnt7Vp06y1qnjs37EgAAAAQYwpQAc457b3+eu356tfU+ud/rnlfuU4N06aFHQsAAAABigJMKlco6Ol/vVr7br1Vx154gU7853+WNTeHHQsAQnHHA/8QdoRQvHHVtWFHADAGigJMGjcwoF1XfVzP//SnmrlmjeZ+7ApZAz3WAAAAqg1FASZF4eBB7fjQh9T7m99qzoc/rFmXXSozCzsWgDL56v98NOwIofjga74UdgQAmBQUBSi7/HPPaft73quDjz+uEz7zGc34uzeEHQkAAACjoChAWQ08vVvbL7lE/U88oXnXXafp/99fhB0JAAAAY6AoQNn0P/GEtl18ifL79mnBv9+g+KpVYUcCAABAESgKUBZ+KqVtl14mFQpa2Nmpae0vDjsSAAAAisRUMJiw3gcf0ta3vV0WjWjRbbdREAAAANQYigJMyIH//V9tv/RSNZ1wghbffruiS04OOxIAAABKRFGAcXvuRz/Sjg98UNFEmxZ971Y1n3BC2JEAAAAwDowpwLjsvfFG7f7ilxTv6ND8r31VDfF42JEAAAAwThQFKIlzTruv/aKevflmTf/L83XS1VfLIpGwYwEAAGACKApQNJfL6clPfFL7f/xjzXjzm3T8xz8ua2wMOxYAAAAmiKIARSn4vnZ+5KPq+dWvNPvyyzX7fe+VmYUdCwAAAGVAUYAx5Q8c0I73vFd9jz6q4z/xT5r5lreEHQkAAABlRFGAUeWeeUbbLr1M2Y0bddK11+rYv/6rsCMBAACgzCgKcFT9O3Zo2zsvVm7PHi341jfVunp12JEAAAAwCSgKMCI/s0HbL7lEhf5+Lbz5JrWceWbYkQAAADBJWLwMR+h79FFtfdvbpIYGLf7erRQEAAAAdY6iAC9w4J57tO3iS9Q0c6YW336bokuXhh0JAAAAk4yiAIfsv/NO7Xjf5YouWaJFt31PzfPmhR0JAAAAFUBRAEnSs52d2nXFx9Ty0pdq4S2dapo1K+xIAAAAqBAGGk9xzjnt+epXtff6b+uY17xGJ33xWjVEo2HHAgAAQAVRFExhLp/XU5/9nJ77/vd13P99vU749KdljY1hxwIAAECFURRMUYX+fu36hyt04Be/0KzLLtOcD39IZhZ2LAAAAISAomAKyvf0asf7L1ff/Q9o7sc+plnvuCjsSAAAAAgRRcEUk3v2WW2/7F3yPU8nXv2vOu61rw07EgAAAEJGUTCFDOzapW0XX6KBXbs0/+tf1zGvfEXYkQAAAFAFKAqmiOymTdp28SUq9PVp4U03quXss8OOBAAAgCpBUTAFHHz8cW2/7F1SpFmLbr1Fsba2sCMBAACgirB4WZ3r6Vqrre94pxqmT9fi226jIAAAAMARKArq2PM//7m2v+c9iixYoEW3fU+RhQvDjgQAAIAqRFFQp569/Xbt/MhHNe2M07Xo1lvUPHdu2JEAAABQpRhTUGecc3rmm9/UM1//N7W+4hWad92X1RCLhR0LAAAAVYyWgnrjnPr/+Ecde+GFmv+1r1IQAAAAYEy0FNQZa2jQSddcIzU2yhqo+QAAADA2ioI6ZM3NYUcAAABADeGjZAAAAGCKoygAAAAApjiKAgAAAGCKoygAAAAApjiKAgAAAGCKoygAAAAApjiKAgAAAGCKoygAAAAAKsTMzjOzjJltMrMrR3j+ZWb2OzPLmdnrK5WLogAAAACoADNrlPQNSedLWi7pTWa2/LDDtkm6SNLtlczGisYAAABAZayUtMk5t1mSzOwOSRdKSg0d4Jx7IniuUMlgtBQAAAAAlTFP0vZh2zuCfaGjpQAAAAAogxmNjU1eW+KRYbtuSKS9G0ILVAKKAgAAAKAM9uXzuUTaO3uUQ3ZKWjBse36wL3R0HwIAAAAq42FJS83sZDOLSHqjpDtDziSJloJJddqnLw47QijWf/qmsCMAAABUHedczswul/QLSY2SbnbOrTezz0p6xDl3p5m9VNKPJc2Q9Ddm9hnn3GmTnY2iAAAAAKgQ59xdku46bN8nhz1+WIPdiiqK7kMAAADAFEdRAAAAAExxNVkUjLU8NAAAAIDi1VxRUOTy0AAAAACKVIsDjcdcHhq17TVfeXfYEULxPx+6PuwIU9JVP/5w2BFC8fm/vS7sCACAKlJzLQWq4uWhAQAAgFpkzrmwM5TEzF4v6Tzn3CXB9tskneOcu3zYMZdJuizYPEvSwYoHDV+TpFzYIWoQ9218uG/jw30bH+7b+HDfxof7Nj5T9b5Nc87V4ofuNdl9aMzloZ1zN0i6oZKhqo2ZPeKcG22ZbYyA+zY+3Lfx4b6ND/dtfLhv48N9Gx/uW+2pxUqmapeHBgAAAGpRzbUUHG156JBjAQAAADWr5ooCaeTloXGEKd19agK4b+PDfRsf7tv4cN/Gh/s2Pty38eG+1ZiaG2gMAAAAoLxqcUwBAAAAgDKiKBiFmfWU4Rxnm9nXRnl+sZm9udjjR/j+e8wsY2aPm9nDZrZiopnLxcwuMLMrS/yevJmtM7P1wd/po2Y2rtepmX3WzF49yvPvNrO3j+fcw87RHuRdZ2bPmtmW4PH/TuS8RVx36D79wcx+ambHlem8F5nZv5XjXIedd+h1OnSvXl/uawTXecHPU5nOecTvgWJeO6PdSzO76rDt483sdjPbbGaPmtn9Zva3wXMvN7P9wX37vZn9r5nNHXYNN/x1bmavDfZNyj0eDzM7wczuMLM/Bn+/u8zs1KMce5yZvXfY9svN7GeTmO0iMztpss4/Xma2IPh9MjPYnhFsLzazpWb2s2H389dm9rLguIvMbM+w36M/NLOWMuZaYWZ/Wa7zlduw13/bUZ6/x8xGnRHnsN9XXjDNeTkzVuVrbjgzmzXs9/VTZrZz2HbksGO/O9Lvm+Bn15nZ3wzb9zMze/kY1676+1OvKAommXPuEefcB0Y5ZLGkQ29iijh+JG9xzp0h6ZuSri095ZHMrHGi53DO3emcu7rEbzvonFvhnDtN0msknS/pU+O8/iedc0d9c+6cu945d8t4zj3sHN1B3hUanAXrH4Lt4W/SJmPsztB9erGkZyW9bxKuUW5vGbpXzrkfFvMN47h3izXs52mylOG1c6goMDOT9F+SfuucW+Kce4kGZ1WbP+z4e4P7droGZ2Ab/u/dHRw/5E2SHp9AtrIK/n4/lnSPc+5Fwd/vHyUdf5RvOU7Se4/y3GS4SFLVvQFxzm2X9C1JQ79Dr9ZgH+2nJP23pBuG3c/3S1oy7Nu/P+z3aL+kvytjtBWSqrYo0ODrvyv4cyLeEvxe75B0zeFvhCfoIlXha24459zeYf+3XS/pumG/v/tLONUOSR8v8fIXqcrvT72iKChR8CnJA8Endj82sxnB/pcG+9aZ2bVm9odg/6FPuczsz4dV2o+Z2TEa/EW/Otj34cOObzWz75hZd3Du140R734FqzubWdzMbjazh4JrXRjsbzGzH5hZKsj/4NCnJmbWY2ZfMrPHJf2Zmb01+P51ZvZtM2sMvr5rg59Qd5vZh4Pv/UBwzt+b2R3BvkOflAafbv0qeP6XZrYw2P9dM/uamd1nZps1OKOUJMk5t1uDi9BdboMag3v7cHCedw37d/lYkOdxM7t62LlfHzy+eli+Lwb7Pm1mfz/Gv+s9ZnZNcB82mNnqIl8n95jZV8zsEUkfNLOXmNlvbPBTvV+Y2YnBcS8ys7uD/ffaUT7dKuHffaUNfsr8WHBPlw37t/jP4FobzewLw7K+I/i7PaTB/wCH9o/2b/at4H5tDl6zN9vgJ2rfLTa0mc00s/8Kzv+AmZ0e7P+0md1qZmsl3Wpmc8zsR8G/+8Nm1hEcN+bP0zjuZbHZh792RvzZD5x0+D0PXp/TguNvk/RKSf3OueuHvsk5t9U59/URrmuSjpG0b9jueyWtNLNmM2uVdIqkdeX+O0/AKyQNHPb3e1zSY8Hr6nfBz+6FwdNXS3rR0P0M9k03s/+2wU9vr7eg9dDM3hR87x/M7Jqh84+030b43RX8fjhb0m3B9aZV4H6U4jpJq84Q814AAA2MSURBVMzsQ5KSkr4o6S2S7nfOHZqK2zn3B+fcdw//ZhssquMKXi+j/Ewfbf//De7X42b2Wxt8Y/xZSX8X3K9yFhsTFrz+k5IuVlAo2//f3plHW11VcfzzRQg0EFOprBDRyDQ1bTAtU8uhuawcIkqxsmxwKktbSwsbbBTJHJpMSLHMNBVdmqhgLSwRHJBBrESxIsXykSKk8XZ/7H19h8v9vXcfvsfjvbc/a911zz3n9zv3zOfss/f5/aRN5VqqRZJ+C2xaXH+BpDlyjcoZFdEOBVYCa+Kevt7mKpF0TIzB98SYXGqgDoyyvF/Suwv/e4AVkg5qEN8682JvLp8+gZnlp+IDPNnAbx6wX7i/BkwK93xg73B/G5gf7v2Ba8M9DXhTuIfiT396NrzB9d+pxR+/X9AgPTOB14X7RODMcJ8JfCTcWwD345PDycCPw38X/G2DtfsNODzcO0V6B8Xv84EjgdcC04v/3yK+/wEMrvMbD5xb5P2ocH8MuCrck4HLcQF1Z6C1QR5b8F3FTwKnhd9gYA4wGtcm3AZsFmFbFnEfCmwFLKbtYH0tfROAkzuo15nAWeF+J3BTO+1lMnBocd/54R4U6RsRv4/AH6ULcDMwJtxvAG7pTNvEhajL8bd8A2wODAz3gcAVRV08AAwHhgAP4S8B3AZYCowAngfMarLOfgUIeB/wH2DXqMO5wO4V7XQxvli9O+rkh8BXI/ytwN1FvczF3woJcCmwT7i3BRY125+6cRwo205V329Y5vVxAsfju3BV/78/sCLK7WHgPmDzso8BE4F34wvGr1K0xZ7+VOUv6quWj62Bv0Sb2q5WhkX+V+M74ZsA0/F+/ZKi7Q4EbgEOace/auyaSYyBG+MHeBs+Nh8UvycCJ7Rz/XhgebSXR3ChcZOizxwV7rJPV/nfC7y0rrzGE2PExvaJ9n9huG+LOv88bePtbqw959Xmik2iHexWtInF+LywCvhU+PeLNtegXCfga4etCr9vAMeFezJwAz4HjMG1A0Oi714L7AvcGtdeG/7tzYu9qnz60ic1BZ1A0nC8U98aXlOAfeX23MPM7I/hf2lFFLOAiZKOj3g6ev33gcB5tR9m9njFdVMlLcFVdLXrDwZOlXQ33sGG4AuqffAFHWY2Hx/0aqwBrgj3AfiAdkfEcQA+KT8AbC/ph5Leji8IiXimSvoIjV9rvjdt5XJxpKPGVWbWamYL8UVBFQcDR0Z6bscXlmPwcrrIzJ6KfP277r4V+KLiQkkfAJ4qA6vqtbjkyvieiy9YmuWy+N4RF8CmR9pPA14Wu1pvBC4P/x/ji/Rm2DTu+ScuME0P/+ER33x8l/FVxT03m9kKM1sNLARG4YLITDNbbq4Svqy4vr06m2Y+et8LPGJuRtUKLKC6jErzoX9FfBcDmNktwFaSNo9rrzGzVeE+EDg38nsNvms8lM73py6nib7fqMw7ivO82Im7o/CumQ+NBC4Cvlt326/wndEPAb9cn7z0AALOlDQPuAnXdlWZFM02swfMbA2ev32A19PWdv8HTMX7bZV/1di1sfMOYBk+hqyDXLM5X9KVhfdl5mYfL8b76BfDv6pPV/nPAiZLOoZCi7sRM5aY3+J7LF73lwCY2TzWnvMOl3QncBc+Vu5chI0zN9fbFjhZ0ij6T5urYhe5RvteXAAr55dfxzz+Zzzfz2q9zez3AJLKOaThvNjdGUjaJ4WCDYi5ff0ncPXlLK2fqUgjxuEL9in47iv4hPvBYhG2rZkt6iCe1THp1u6fUty/o5lNCMHk1bigcSzws7j+XbhA8hpckOiMLfh/qwIkbY8LK49Gmo4r0jTazG7sKPIYpPcEfoPvpt7QibSV6VtD597tsTK+BSwo0r2rmR2M97+Wwn93M9upybhXxaQ/KuKv2Zh/HZhhftbgPbgwWJ+P9clLPbW4WuvibX2O8dZYWbgHAHsVZfRSM3uyG/tTV9JMmS/A+w0AZvZZXAgfURHnNawttGJms3FtzdZmdv9zSXA3sADfYKhnHJ7H10ZbfoS122tJ/bOzO/0s7XbGro0W+YMjDgL2Ak6Smx3Wt5f347v3W9bfH4L7NOraS7OY2bH4Ym0kMFfSVusTz4ZAfiD7rcDPJD2IC0KHU7HRJGk0vvt9QCz+r6NB+zOz5cCd+AZKp+iNba4DJgOfM7NdgTNYu7w66qPfxNtSjap5MelBUijoBGa2AnhcbXblH8VVYi3AE5Jqg8aHGt0vaYfYUf0OfljwlcATuI1wI6ZTHChU2LlXpM2A03H701fib3w+TpLi3j3i0ln4QImknfGFRCNuBg5V21NOtpQ0StLWwAAzuwLv4K+R2/eONLMZwCn4bvXQuvhuo61cxuEq7XaRNAI/4HRu5O93wKclDYrwV0h6Pl5OR9fsG2NyKOMZCgw3f+ndSfgg/SxV9dpR+jrBYmCEpL0jPYMkvcrM/gMskXRY+EvSq9uLqJ7QjhwPfCEEseHA3yN4fBNR3A7sJ3/SxCDgsCKs03XWSf4Q8SJ/GsVjUSb13IgfpCSu3T2+O9ufupxm+34Dnqm1Y9zUYIikTxfh7T0tZh/grw38T6U4wLwRcQswWMUTXOTnR0YBj5rZM5LeQpsWpVEd7ilpdIw1R+AHSWfjbXdr+YMRxuL9tqF/o7Grnf/rcWLsvgA40cyW4g+R+D6+o/8mSe8tLm+2vVT16Yb+0cduN7Ov4CZJI9lIyws3KbvYzEaZ2XahVVuCa3g/DCBpF9yECNzUciVu7/4iXCOzDjGv7IGXYZ9uc00wDFgWY9e4urDDJA2QtAO+Sbm4DIwNvBfQVv4N58UI663l0+vplW803oBsJulvxe+JwFHAj2KgeAA4OsI+DvxUUis+SKxoEN+JMfnVzCyuD/ca+eHeybgas8Y3gPPCFGQNLplfSQVmtkrSWfgOyeeAScC8mEiX4Lvk5wNTJC3EbZMXNEqrmS2UdBpwY9z/DC6grAIuUttjQr+Mq5UvCTMcAeeYWUvIIzWOi/u+iE8uR1NBqBIH4WZIF+PlDr7Lsh1wZ0yYy4FDzOyGWCjOkfQ0/rbrcnE0DLha0pBI3+cb/G1VvT5nzOxp+eGpc6KMBuJ1swAfWC+Ish6Eq7w79eQYM7srTDDG4mYlUyK+65q4d5mkCfhh5RbWPqDadJ2tJxOAn0fan8LroBHH4/1gHl52v8d33TrsT2Z2dheks9E4UNJM36/nJ3jfvNPMxkk6BDhb0pfwsl6JC9g13hz9QhH/J+ojNLPrm87RBsTMTP541UmSTsFN+R7E6/+cMEWYg49HmNm/JM2Kce96vB3fgZ+deDkwA/itmbXKH3k8Ay+X68zsaoBG/iFw149d4OPujyStws+G1MzWeppjgKVmVjMNPB/vg3viY/lESZNwDcsT+HxR4wi5qcYA3L57fPhX9ekq/+9JGoOX48342LSUNtPUb5lZaXLYk4zFz+GVXIEv6DeVtAhYhAsJmNk9ku7C293D+IZZydRoE4PxsWQuNG5bVf69sM11xOn4RtLy+C4X7ktx4Whz4FgzW123BgDXFlwNHc6Lk+md5dPryTcadxGShprZk+E+FdjGzE7o4WStQ+xiDIoOuwNuy7ujde4RY0mSBL2l7ydJkiRJe6SmoOt4l6Qv42X6EM2ZbvQEmwEzQv0n4DMpECTJc6K39P0kSZIkqSQ1BUmSJEmSJEnSz8mDxkmSJEmSJEnSz0mhIEmSJEmSJEn6OSkUJEmSJEmSJEk/J4WCJEmSpCkk/S/eKdHMtftL2uBvmU6SJEnWjxQKkiRJ+hCSZkoySYfX+b8h/B/soaQlSZIkGzEpFCRJkvQ9FuEvvyo5JvyTJEmSZB1SKEiSJOl7XAnsIWl7AEnDgA8CF9UukLSZpB9IeljSY5KukrRtET5M0hRJ/5b0kKR13jgt6RBJcyW1SFokadwGyFuSJEnSDaRQkCRJ0vdYDUwFPh6/xwK3AsuKa84G9orPKOAxYFq89RxgEjAG2BnYDXgfUAtD0kHAhcCJwJbAUcC5kvbtniwlSZIk3UkKBUmSJH2TnwJHSxoIfDJ+AyBpAL6IP83M/m5mK/HF/U7AnhE+DjjdzP5pZiuAU+riPwH4gZn9wcxazWw2cAlwZLfnLEmSJOlyUihIkiTpg5jZfOAh4HTghcANRfAIYDCwpLj+SeBRYGQR/mBxzxLWZjRwSpgOtUhqAcYDL+nSjCRJkiQbhBQKkiRJ+i4/wYWCn5vZmsJ/OfBfYLuah6ShuPDwMG5K9HQZXucGFzgmmNkWxWeYmb2zqzORJEmSdD8DezoBSZIkSbfxS3yRP7f0NLNWSb8Avi5pIdACnAXcB8w2szWSLgXOkDQfWAV8uy7uScBkSX8CbsPPG+wKyMzmdGemkiRJkq4nNQVJkiR9FDNbbWY3mdnjDYJPAuYAdwBLgW2A9xYahRNwk6H7gHuBacCz2gYzuxF/zOn3cM3CMvzw8tDuyU2SJEnSncjMejoNSZIkSZIkSZL0IKkpSJIkSZIkSZJ+TgoFSZIkSZIkSdLPSaEgSZIkSZIkSfo5KRQkSZIkSZIkST8nhYIkSZIkSZIk6eekUJAkSZIkSZIk/ZwUCpIkSZIkSZKkn5NCQZIkSZIkSZL0c1IoSJIkSZIkSZJ+zv8By43fuLaFf4MAAAAASUVORK5CYII=\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["The Random Forest model gives the best `f1_score` results but the training time is long and computationally expensive and TabNet model is overfitting, so for optimization we will choose the `LightGBM` model for fine-tuning."],"metadata":{"id":"GKYoPW-MGBXH"}},{"cell_type":"markdown","source":["##Fine Tuning"],"metadata":{"id":"sSrpjXmX5u7u"}},{"cell_type":"code","source":["!pip install --quiet optuna"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"ktwsdWLg69rt","outputId":"cbca120b-bfc1-4985-86b3-baa11891ef1b","executionInfo":{"status":"ok","timestamp":1659227883209,"user_tz":-420,"elapsed":6884,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":51,"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[K     |████████████████████████████████| 308 kB 7.4 MB/s \n","\u001b[K     |████████████████████████████████| 209 kB 82.1 MB/s \n","\u001b[K     |████████████████████████████████| 81 kB 11.1 MB/s \n","\u001b[K     |████████████████████████████████| 78 kB 8.3 MB/s \n","\u001b[K     |████████████████████████████████| 147 kB 90.8 MB/s \n","\u001b[K     |████████████████████████████████| 112 kB 74.7 MB/s \n","\u001b[K     |████████████████████████████████| 49 kB 7.2 MB/s \n","\u001b[?25h  Building wheel for pyperclip (setup.py) ... \u001b[?25l\u001b[?25hdone\n"]}]},{"cell_type":"code","source":["import optuna  # pip install optuna\n","import lightgbm as lgb\n","import xgboost as xgb\n","from sklearn.metrics import make_scorer, accuracy_score, precision_score, recall_score, f1_score\n","\n","def objective(trial, X_train_smt, y_train_smt):\n","    max_depth=trial.suggest_int('max_depth',2,14)\n","    param_grid = {\n","        #         \"device_type\": trial.suggest_categorical(\"device_type\", ['gpu']),\n","        #\"n_estimators\": trial.suggest_categorical(\"n_estimators\", 1,300 ),\n","        \"learning_rate\": trial.suggest_float(\"learning_rate\", 0.0001, 0.3),\n","        \"num_leaves\": trial.suggest_int(\"num_leaves\", 4, 2**max_depth),\n","        \"min_data_in_leaf\": trial.suggest_int(\"min_data_in_leaf\", 5, 500),\n","        \"max_bin\": trial.suggest_int(\"max_bin\", 10, 300),\n","        \"lambda_l1\": trial.suggest_int(\"lambda_l1\", 0, 100, step=5),\n","        \"lambda_l2\": trial.suggest_int(\"lambda_l2\", 0, 100, step=5),\n","        \"min_gain_to_split\": trial.suggest_float(\"min_gain_to_split\", 0, 15),\n","        \"bagging_fraction\": trial.suggest_float(\n","            \"bagging_fraction\", 0.2, 0.95, step=0.1\n","        ),\n","        \"bagging_freq\": trial.suggest_categorical(\"bagging_freq\", [1]),\n","        \"feature_fraction\": trial.suggest_float(\"feature_fraction\", 0.2, 0.95, step=0.1)\n","    }\n","    #scoring = {'accuracy' : make_scorer(accuracy_score), \n","    #      'precision' : make_scorer(precision_score),\n","    #     'recall' : make_scorer(recall_score), \n","    #    'f1_score' : make_scorer(f1_score)}\n","\n","    cv = StratifiedKFold(n_splits=5, shuffle=True, random_state=42)\n","    model = lgb.LGBMClassifier(**param_grid)\n","    model.fit(X_train_smt,y_train_smt)\n","    scores = cross_val_score(model, X_train_smt, y_train_smt,\n","                             scoring=make_scorer(f1_score, average='weighted', labels=[1]),cv=cv, n_jobs=-1)\n","    return np.mean(scores)"],"metadata":{"id":"Ao6_agKnEk0J","executionInfo":{"status":"ok","timestamp":1659227884304,"user_tz":-420,"elapsed":641,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":52,"outputs":[]},{"cell_type":"code","source":["study = optuna.create_study(direction='maximize')\n","func = lambda trial: objective(trial, X_train_smt, y_train_smt)\n","study.optimize(func, n_trials=100)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"L71YMUazFtuj","outputId":"c800cd28-2b84-4386-89eb-b3c96baa1b52","executionInfo":{"status":"ok","timestamp":1659230346767,"user_tz":-420,"elapsed":2462477,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":53,"outputs":[{"output_type":"stream","name":"stderr","text":["\u001b[32m[I 2022-07-31 00:38:04,267]\u001b[0m A new study created in memory with name: no-name-ac0a55cb-d40a-40db-a49a-098168374496\u001b[0m\n","\u001b[32m[I 2022-07-31 00:38:33,969]\u001b[0m Trial 0 finished with value: 0.9974027037921788 and parameters: {'max_depth': 6, 'learning_rate': 0.0808499324423164, 'num_leaves': 38, 'min_data_in_leaf': 261, 'max_bin': 78, 'lambda_l1': 45, 'lambda_l2': 45, 'min_gain_to_split': 7.23178728515051, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 0 with value: 0.9974027037921788.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:38:52,259]\u001b[0m Trial 1 finished with value: 0.9818466886898914 and parameters: {'max_depth': 14, 'learning_rate': 0.19683326680434887, 'num_leaves': 8507, 'min_data_in_leaf': 337, 'max_bin': 110, 'lambda_l1': 85, 'lambda_l2': 65, 'min_gain_to_split': 6.645617046438769, 'bagging_fraction': 0.2, 'bagging_freq': 1, 'feature_fraction': 0.30000000000000004}. Best is trial 0 with value: 0.9974027037921788.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:39:10,555]\u001b[0m Trial 2 finished with value: 0.9950469488328586 and parameters: {'max_depth': 6, 'learning_rate': 0.2914902829211896, 'num_leaves': 26, 'min_data_in_leaf': 167, 'max_bin': 190, 'lambda_l1': 55, 'lambda_l2': 85, 'min_gain_to_split': 12.582643936879892, 'bagging_fraction': 0.5, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 0 with value: 0.9974027037921788.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:39:29,972]\u001b[0m Trial 3 finished with value: 0.969627419516244 and parameters: {'max_depth': 6, 'learning_rate': 0.031928437771218156, 'num_leaves': 61, 'min_data_in_leaf': 404, 'max_bin': 278, 'lambda_l1': 85, 'lambda_l2': 100, 'min_gain_to_split': 12.425372139838295, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.2}. Best is trial 0 with value: 0.9974027037921788.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:39:51,334]\u001b[0m Trial 4 finished with value: 0.9685449136404071 and parameters: {'max_depth': 2, 'learning_rate': 0.16759084857856096, 'num_leaves': 4, 'min_data_in_leaf': 194, 'max_bin': 10, 'lambda_l1': 80, 'lambda_l2': 100, 'min_gain_to_split': 5.551608249125443, 'bagging_fraction': 0.5, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 0 with value: 0.9974027037921788.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:40:16,719]\u001b[0m Trial 5 finished with value: 0.9974710098625323 and parameters: {'max_depth': 11, 'learning_rate': 0.1227550509731993, 'num_leaves': 468, 'min_data_in_leaf': 314, 'max_bin': 65, 'lambda_l1': 75, 'lambda_l2': 65, 'min_gain_to_split': 6.6920543511862105, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 5 with value: 0.9974710098625323.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:40:33,479]\u001b[0m Trial 6 finished with value: 0.9956460578521668 and parameters: {'max_depth': 7, 'learning_rate': 0.24926014350378203, 'num_leaves': 69, 'min_data_in_leaf': 489, 'max_bin': 67, 'lambda_l1': 20, 'lambda_l2': 90, 'min_gain_to_split': 11.54790834654213, 'bagging_fraction': 0.4, 'bagging_freq': 1, 'feature_fraction': 0.4}. Best is trial 5 with value: 0.9974710098625323.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:40:54,389]\u001b[0m Trial 7 finished with value: 0.9809227958027563 and parameters: {'max_depth': 9, 'learning_rate': 0.003296367293270812, 'num_leaves': 142, 'min_data_in_leaf': 116, 'max_bin': 66, 'lambda_l1': 5, 'lambda_l2': 20, 'min_gain_to_split': 12.812210739932683, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.30000000000000004}. Best is trial 5 with value: 0.9974710098625323.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:41:07,251]\u001b[0m Trial 8 finished with value: 0.981075390839783 and parameters: {'max_depth': 4, 'learning_rate': 0.24799517687279526, 'num_leaves': 13, 'min_data_in_leaf': 473, 'max_bin': 102, 'lambda_l1': 70, 'lambda_l2': 20, 'min_gain_to_split': 14.125931414075538, 'bagging_fraction': 0.2, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 5 with value: 0.9974710098625323.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:41:23,331]\u001b[0m Trial 9 finished with value: 0.9928429694992958 and parameters: {'max_depth': 11, 'learning_rate': 0.258052587352991, 'num_leaves': 368, 'min_data_in_leaf': 443, 'max_bin': 48, 'lambda_l1': 70, 'lambda_l2': 75, 'min_gain_to_split': 10.233850652537868, 'bagging_fraction': 0.4, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 5 with value: 0.9974710098625323.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:41:56,336]\u001b[0m Trial 10 finished with value: 0.999160858032751 and parameters: {'max_depth': 13, 'learning_rate': 0.10561411944207447, 'num_leaves': 2515, 'min_data_in_leaf': 24, 'max_bin': 190, 'lambda_l1': 40, 'lambda_l2': 45, 'min_gain_to_split': 1.0879712022318015, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 10 with value: 0.999160858032751.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:42:29,816]\u001b[0m Trial 11 finished with value: 0.9991060182038065 and parameters: {'max_depth': 13, 'learning_rate': 0.09893261946542006, 'num_leaves': 2754, 'min_data_in_leaf': 50, 'max_bin': 187, 'lambda_l1': 40, 'lambda_l2': 40, 'min_gain_to_split': 0.674023406292326, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 10 with value: 0.999160858032751.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:43:03,665]\u001b[0m Trial 12 finished with value: 0.9992487169730648 and parameters: {'max_depth': 14, 'learning_rate': 0.0968170877572363, 'num_leaves': 5141, 'min_data_in_leaf': 9, 'max_bin': 200, 'lambda_l1': 30, 'lambda_l2': 40, 'min_gain_to_split': 0.8762499939741106, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 12 with value: 0.9992487169730648.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:43:45,786]\u001b[0m Trial 13 finished with value: 0.998984913649306 and parameters: {'max_depth': 12, 'learning_rate': 0.055767005797778574, 'num_leaves': 3754, 'min_data_in_leaf': 5, 'max_bin': 257, 'lambda_l1': 25, 'lambda_l2': 0, 'min_gain_to_split': 0.8902995864172247, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 12 with value: 0.9992487169730648.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:44:17,980]\u001b[0m Trial 14 finished with value: 0.9989984270824923 and parameters: {'max_depth': 14, 'learning_rate': 0.1485867864300197, 'num_leaves': 12216, 'min_data_in_leaf': 75, 'max_bin': 228, 'lambda_l1': 30, 'lambda_l2': 30, 'min_gain_to_split': 3.0410192141263046, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 12 with value: 0.9992487169730648.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:44:48,321]\u001b[0m Trial 15 finished with value: 0.9994507466455584 and parameters: {'max_depth': 9, 'learning_rate': 0.11445479956180926, 'num_leaves': 418, 'min_data_in_leaf': 6, 'max_bin': 155, 'lambda_l1': 0, 'lambda_l2': 55, 'min_gain_to_split': 3.1562491704850144, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:45:11,815]\u001b[0m Trial 16 finished with value: 0.9993651035377689 and parameters: {'max_depth': 9, 'learning_rate': 0.1848703394784552, 'num_leaves': 217, 'min_data_in_leaf': 117, 'max_bin': 146, 'lambda_l1': 0, 'lambda_l2': 60, 'min_gain_to_split': 3.400153746916481, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:45:33,788]\u001b[0m Trial 17 finished with value: 0.9993233969213113 and parameters: {'max_depth': 9, 'learning_rate': 0.21342382139854424, 'num_leaves': 312, 'min_data_in_leaf': 111, 'max_bin': 139, 'lambda_l1': 0, 'lambda_l2': 60, 'min_gain_to_split': 4.0165779274855, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:45:54,747]\u001b[0m Trial 18 finished with value: 0.997340830510239 and parameters: {'max_depth': 9, 'learning_rate': 0.17220916603821498, 'num_leaves': 284, 'min_data_in_leaf': 159, 'max_bin': 147, 'lambda_l1': 100, 'lambda_l2': 55, 'min_gain_to_split': 3.645939513843402, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:46:16,138]\u001b[0m Trial 19 finished with value: 0.9987701965949423 and parameters: {'max_depth': 8, 'learning_rate': 0.13120546290863147, 'num_leaves': 250, 'min_data_in_leaf': 220, 'max_bin': 123, 'lambda_l1': 10, 'lambda_l2': 75, 'min_gain_to_split': 9.532290845388767, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:46:38,570]\u001b[0m Trial 20 finished with value: 0.9991148526912811 and parameters: {'max_depth': 10, 'learning_rate': 0.20883665024247672, 'num_leaves': 290, 'min_data_in_leaf': 96, 'max_bin': 168, 'lambda_l1': 15, 'lambda_l2': 75, 'min_gain_to_split': 2.767175614598899, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:46:59,959]\u001b[0m Trial 21 finished with value: 0.9992553215006181 and parameters: {'max_depth': 8, 'learning_rate': 0.20776456449864766, 'num_leaves': 253, 'min_data_in_leaf': 129, 'max_bin': 135, 'lambda_l1': 0, 'lambda_l2': 60, 'min_gain_to_split': 5.324264702920854, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:47:23,972]\u001b[0m Trial 22 finished with value: 0.9993914577106089 and parameters: {'max_depth': 10, 'learning_rate': 0.1794144438485037, 'num_leaves': 347, 'min_data_in_leaf': 58, 'max_bin': 157, 'lambda_l1': 0, 'lambda_l2': 50, 'min_gain_to_split': 4.329874369356489, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:47:49,574]\u001b[0m Trial 23 finished with value: 0.9993848583949427 and parameters: {'max_depth': 10, 'learning_rate': 0.17532372064902144, 'num_leaves': 341, 'min_data_in_leaf': 50, 'max_bin': 161, 'lambda_l1': 10, 'lambda_l2': 30, 'min_gain_to_split': 2.2678107026734766, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:48:17,265]\u001b[0m Trial 24 finished with value: 0.9993892488822128 and parameters: {'max_depth': 11, 'learning_rate': 0.14471577365507476, 'num_leaves': 627, 'min_data_in_leaf': 58, 'max_bin': 218, 'lambda_l1': 10, 'lambda_l2': 30, 'min_gain_to_split': 2.1490151686521406, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:48:41,438]\u001b[0m Trial 25 finished with value: 0.999101597895413 and parameters: {'max_depth': 11, 'learning_rate': 0.14413640545171277, 'num_leaves': 672, 'min_data_in_leaf': 64, 'max_bin': 227, 'lambda_l1': 20, 'lambda_l2': 0, 'min_gain_to_split': 5.456142928969686, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:49:02,317]\u001b[0m Trial 26 finished with value: 0.9988051751766625 and parameters: {'max_depth': 12, 'learning_rate': 0.12384279208856783, 'num_leaves': 922, 'min_data_in_leaf': 41, 'max_bin': 218, 'lambda_l1': 10, 'lambda_l2': 30, 'min_gain_to_split': 8.568566261196342, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:49:34,286]\u001b[0m Trial 27 finished with value: 0.9980120457596149 and parameters: {'max_depth': 10, 'learning_rate': 0.067435254304688, 'num_leaves': 460, 'min_data_in_leaf': 83, 'max_bin': 293, 'lambda_l1': 55, 'lambda_l2': 20, 'min_gain_to_split': 4.7360937617163685, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:50:01,137]\u001b[0m Trial 28 finished with value: 0.999329985186718 and parameters: {'max_depth': 7, 'learning_rate': 0.16028034363028051, 'num_leaves': 128, 'min_data_in_leaf': 154, 'max_bin': 242, 'lambda_l1': 15, 'lambda_l2': 50, 'min_gain_to_split': 1.778810622579411, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:50:27,741]\u001b[0m Trial 29 finished with value: 0.9989523534489468 and parameters: {'max_depth': 12, 'learning_rate': 0.07239715817592099, 'num_leaves': 1147, 'min_data_in_leaf': 244, 'max_bin': 164, 'lambda_l1': 5, 'lambda_l2': 40, 'min_gain_to_split': 0.11204461165383783, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:50:45,557]\u001b[0m Trial 30 finished with value: 0.958889888646036 and parameters: {'max_depth': 5, 'learning_rate': 0.04362547580521674, 'num_leaves': 6, 'min_data_in_leaf': 290, 'max_bin': 101, 'lambda_l1': 30, 'lambda_l2': 10, 'min_gain_to_split': 7.745409034879571, 'bagging_fraction': 0.30000000000000004, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:51:08,829]\u001b[0m Trial 31 finished with value: 0.9993497323656779 and parameters: {'max_depth': 10, 'learning_rate': 0.22919962517536585, 'num_leaves': 404, 'min_data_in_leaf': 47, 'max_bin': 170, 'lambda_l1': 10, 'lambda_l2': 30, 'min_gain_to_split': 2.0483808951482625, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:51:34,861]\u001b[0m Trial 32 finished with value: 0.9994178057533606 and parameters: {'max_depth': 10, 'learning_rate': 0.1856873593152291, 'num_leaves': 440, 'min_data_in_leaf': 66, 'max_bin': 209, 'lambda_l1': 5, 'lambda_l2': 50, 'min_gain_to_split': 2.5235803735221225, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:51:58,511]\u001b[0m Trial 33 finished with value: 0.9993146013227342 and parameters: {'max_depth': 11, 'learning_rate': 0.19470101477957819, 'num_leaves': 668, 'min_data_in_leaf': 75, 'max_bin': 212, 'lambda_l1': 0, 'lambda_l2': 50, 'min_gain_to_split': 4.393351851283209, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:52:28,451]\u001b[0m Trial 34 finished with value: 0.999393661474293 and parameters: {'max_depth': 7, 'learning_rate': 0.15191749837840118, 'num_leaves': 127, 'min_data_in_leaf': 354, 'max_bin': 269, 'lambda_l1': 5, 'lambda_l2': 50, 'min_gain_to_split': 2.1732494320031375, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:52:55,704]\u001b[0m Trial 35 finished with value: 0.9990204068814046 and parameters: {'max_depth': 8, 'learning_rate': 0.11441899266116679, 'num_leaves': 256, 'min_data_in_leaf': 375, 'max_bin': 261, 'lambda_l1': 20, 'lambda_l2': 50, 'min_gain_to_split': 4.738432955009455, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:53:23,443]\u001b[0m Trial 36 finished with value: 0.9993453496500388 and parameters: {'max_depth': 7, 'learning_rate': 0.18722097055738834, 'num_leaves': 106, 'min_data_in_leaf': 356, 'max_bin': 298, 'lambda_l1': 5, 'lambda_l2': 70, 'min_gain_to_split': 2.9101233530880384, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:53:46,414]\u001b[0m Trial 37 finished with value: 0.9981582612229886 and parameters: {'max_depth': 5, 'learning_rate': 0.15803952527685522, 'num_leaves': 28, 'min_data_in_leaf': 189, 'max_bin': 268, 'lambda_l1': 15, 'lambda_l2': 45, 'min_gain_to_split': 6.2776962774332095, 'bagging_fraction': 0.5, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:54:07,547]\u001b[0m Trial 38 finished with value: 0.9983576112273902 and parameters: {'max_depth': 7, 'learning_rate': 0.2988151088455572, 'num_leaves': 112, 'min_data_in_leaf': 289, 'max_bin': 244, 'lambda_l1': 55, 'lambda_l2': 65, 'min_gain_to_split': 3.8407186202552817, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:54:26,759]\u001b[0m Trial 39 finished with value: 0.9817849859113998 and parameters: {'max_depth': 3, 'learning_rate': 0.1352934658544205, 'num_leaves': 5, 'min_data_in_leaf': 139, 'max_bin': 179, 'lambda_l1': 25, 'lambda_l2': 85, 'min_gain_to_split': 6.2568631344952506, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:54:46,542]\u001b[0m Trial 40 finished with value: 0.9981718698054831 and parameters: {'max_depth': 6, 'learning_rate': 0.08698314647347989, 'num_leaves': 64, 'min_data_in_leaf': 404, 'max_bin': 89, 'lambda_l1': 5, 'lambda_l2': 55, 'min_gain_to_split': 1.580501313110203, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.2}. Best is trial 15 with value: 0.9994507466455584.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:55:13,277]\u001b[0m Trial 41 finished with value: 0.9995276076144746 and parameters: {'max_depth': 9, 'learning_rate': 0.15341203396419936, 'num_leaves': 415, 'min_data_in_leaf': 31, 'max_bin': 206, 'lambda_l1': 0, 'lambda_l2': 35, 'min_gain_to_split': 2.521180943498521, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 41 with value: 0.9995276076144746.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:55:38,751]\u001b[0m Trial 42 finished with value: 0.99953419756847 and parameters: {'max_depth': 9, 'learning_rate': 0.17184161865253794, 'num_leaves': 413, 'min_data_in_leaf': 26, 'max_bin': 200, 'lambda_l1': 0, 'lambda_l2': 35, 'min_gain_to_split': 2.913913367010989, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 42 with value: 0.99953419756847.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:56:10,367]\u001b[0m Trial 43 finished with value: 0.999740677548943 and parameters: {'max_depth': 9, 'learning_rate': 0.22944980784811914, 'num_leaves': 418, 'min_data_in_leaf': 13, 'max_bin': 247, 'lambda_l1': 0, 'lambda_l2': 35, 'min_gain_to_split': 0.09848343667603521, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:56:35,165]\u001b[0m Trial 44 finished with value: 0.9996989357996275 and parameters: {'max_depth': 9, 'learning_rate': 0.2767446267547524, 'num_leaves': 430, 'min_data_in_leaf': 15, 'max_bin': 199, 'lambda_l1': 0, 'lambda_l2': 35, 'min_gain_to_split': 0.6069391074201935, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:57:02,184]\u001b[0m Trial 45 finished with value: 0.9997187078336826 and parameters: {'max_depth': 9, 'learning_rate': 0.27246222217751426, 'num_leaves': 410, 'min_data_in_leaf': 21, 'max_bin': 201, 'lambda_l1': 0, 'lambda_l2': 35, 'min_gain_to_split': 0.2298007549667247, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:57:25,050]\u001b[0m Trial 46 finished with value: 0.9994924739661034 and parameters: {'max_depth': 8, 'learning_rate': 0.28045793218025256, 'num_leaves': 256, 'min_data_in_leaf': 37, 'max_bin': 194, 'lambda_l1': 15, 'lambda_l2': 35, 'min_gain_to_split': 0.026569079250856387, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.4}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:57:44,821]\u001b[0m Trial 47 finished with value: 0.9992575132467048 and parameters: {'max_depth': 9, 'learning_rate': 0.27586555743626157, 'num_leaves': 418, 'min_data_in_leaf': 23, 'max_bin': 239, 'lambda_l1': 20, 'lambda_l2': 20, 'min_gain_to_split': 1.0514732862383305, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:58:04,909]\u001b[0m Trial 48 finished with value: 0.9987283963041784 and parameters: {'max_depth': 8, 'learning_rate': 0.2366685767894369, 'num_leaves': 256, 'min_data_in_leaf': 29, 'max_bin': 181, 'lambda_l1': 65, 'lambda_l2': 25, 'min_gain_to_split': 0.47966380780759676, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:58:24,755]\u001b[0m Trial 49 finished with value: 0.9972044959003863 and parameters: {'max_depth': 9, 'learning_rate': 0.2633424026483712, 'num_leaves': 432, 'min_data_in_leaf': 99, 'max_bin': 199, 'lambda_l1': 90, 'lambda_l2': 15, 'min_gain_to_split': 1.4618213895421264, 'bagging_fraction': 0.5, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:58:50,481]\u001b[0m Trial 50 finished with value: 0.9996418193093699 and parameters: {'max_depth': 6, 'learning_rate': 0.28004473581264494, 'num_leaves': 47, 'min_data_in_leaf': 17, 'max_bin': 232, 'lambda_l1': 0, 'lambda_l2': 35, 'min_gain_to_split': 0.4148082551787624, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:59:16,505]\u001b[0m Trial 51 finished with value: 0.9996396195027446 and parameters: {'max_depth': 6, 'learning_rate': 0.27059055487238437, 'num_leaves': 52, 'min_data_in_leaf': 5, 'max_bin': 227, 'lambda_l1': 0, 'lambda_l2': 35, 'min_gain_to_split': 0.40101661904087477, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 00:59:43,861]\u001b[0m Trial 52 finished with value: 0.9996000854283869 and parameters: {'max_depth': 6, 'learning_rate': 0.2823183445614125, 'num_leaves': 47, 'min_data_in_leaf': 9, 'max_bin': 231, 'lambda_l1': 5, 'lambda_l2': 35, 'min_gain_to_split': 0.022973552690428, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:00:06,653]\u001b[0m Trial 53 finished with value: 0.9994683137612078 and parameters: {'max_depth': 6, 'learning_rate': 0.28407717217403355, 'num_leaves': 49, 'min_data_in_leaf': 5, 'max_bin': 230, 'lambda_l1': 5, 'lambda_l2': 40, 'min_gain_to_split': 0.449616267958602, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:00:28,638]\u001b[0m Trial 54 finished with value: 0.9992729010731514 and parameters: {'max_depth': 5, 'learning_rate': 0.2624561642168556, 'num_leaves': 32, 'min_data_in_leaf': 90, 'max_bin': 247, 'lambda_l1': 10, 'lambda_l2': 25, 'min_gain_to_split': 1.3125251281376744, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:00:53,961]\u001b[0m Trial 55 finished with value: 0.9992860851801467 and parameters: {'max_depth': 4, 'learning_rate': 0.27185166100094255, 'num_leaves': 16, 'min_data_in_leaf': 18, 'max_bin': 256, 'lambda_l1': 0, 'lambda_l2': 45, 'min_gain_to_split': 0.6236343106625557, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:01:14,680]\u001b[0m Trial 56 finished with value: 0.9991456029538176 and parameters: {'max_depth': 4, 'learning_rate': 0.24684266942197153, 'num_leaves': 16, 'min_data_in_leaf': 10, 'max_bin': 230, 'lambda_l1': 5, 'lambda_l2': 25, 'min_gain_to_split': 0.029091161696008463, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.4}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:01:33,829]\u001b[0m Trial 57 finished with value: 0.9987877121159769 and parameters: {'max_depth': 6, 'learning_rate': 0.29028985597323365, 'num_leaves': 48, 'min_data_in_leaf': 42, 'max_bin': 282, 'lambda_l1': 35, 'lambda_l2': 35, 'min_gain_to_split': 1.0933514437275176, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:01:57,843]\u001b[0m Trial 58 finished with value: 0.9993321833987899 and parameters: {'max_depth': 5, 'learning_rate': 0.23013661773315855, 'num_leaves': 32, 'min_data_in_leaf': 106, 'max_bin': 219, 'lambda_l1': 10, 'lambda_l2': 40, 'min_gain_to_split': 0.6334783106768791, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:02:13,603]\u001b[0m Trial 59 finished with value: 0.997586286123832 and parameters: {'max_depth': 6, 'learning_rate': 0.2530799959614861, 'num_leaves': 49, 'min_data_in_leaf': 79, 'max_bin': 252, 'lambda_l1': 15, 'lambda_l2': 35, 'min_gain_to_split': 14.94506074261593, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.4}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:02:33,026]\u001b[0m Trial 60 finished with value: 0.9992179990952179 and parameters: {'max_depth': 8, 'learning_rate': 0.2688761187237285, 'num_leaves': 253, 'min_data_in_leaf': 57, 'max_bin': 235, 'lambda_l1': 25, 'lambda_l2': 10, 'min_gain_to_split': 1.3678804577276071, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:02:56,940]\u001b[0m Trial 61 finished with value: 0.9995825132955742 and parameters: {'max_depth': 7, 'learning_rate': 0.29689739642841984, 'num_leaves': 56, 'min_data_in_leaf': 25, 'max_bin': 25, 'lambda_l1': 0, 'lambda_l2': 35, 'min_gain_to_split': 0.4464885607275153, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:03:19,866]\u001b[0m Trial 62 finished with value: 0.9994968614121404 and parameters: {'max_depth': 7, 'learning_rate': 0.2876813790305702, 'num_leaves': 53, 'min_data_in_leaf': 21, 'max_bin': 19, 'lambda_l1': 0, 'lambda_l2': 45, 'min_gain_to_split': 0.408276180937557, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:03:39,135]\u001b[0m Trial 63 finished with value: 0.9992838810309044 and parameters: {'max_depth': 6, 'learning_rate': 0.29380907225831276, 'num_leaves': 46, 'min_data_in_leaf': 40, 'max_bin': 48, 'lambda_l1': 5, 'lambda_l2': 25, 'min_gain_to_split': 1.7787912529539958, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:03:57,482]\u001b[0m Trial 64 finished with value: 0.9988492782055193 and parameters: {'max_depth': 7, 'learning_rate': 0.23789013258249275, 'num_leaves': 54, 'min_data_in_leaf': 13, 'max_bin': 223, 'lambda_l1': 0, 'lambda_l2': 40, 'min_gain_to_split': 11.135058998209047, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:04:18,569]\u001b[0m Trial 65 finished with value: 0.9995100320544225 and parameters: {'max_depth': 8, 'learning_rate': 0.2795133498676238, 'num_leaves': 256, 'min_data_in_leaf': 5, 'max_bin': 186, 'lambda_l1': 10, 'lambda_l2': 30, 'min_gain_to_split': 0.801588135487638, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:04:45,271]\u001b[0m Trial 66 finished with value: 0.999441962631936 and parameters: {'max_depth': 6, 'learning_rate': 0.2987236714127614, 'num_leaves': 53, 'min_data_in_leaf': 68, 'max_bin': 211, 'lambda_l1': 5, 'lambda_l2': 35, 'min_gain_to_split': 0.07088030118395841, 'bagging_fraction': 0.4, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:05:07,986]\u001b[0m Trial 67 finished with value: 0.9985290243109703 and parameters: {'max_depth': 5, 'learning_rate': 0.21779584497874196, 'num_leaves': 32, 'min_data_in_leaf': 51, 'max_bin': 273, 'lambda_l1': 50, 'lambda_l2': 25, 'min_gain_to_split': 0.9728401696355412, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:05:34,143]\u001b[0m Trial 68 finished with value: 0.9992312035180853 and parameters: {'max_depth': 4, 'learning_rate': 0.2561419396662445, 'num_leaves': 16, 'min_data_in_leaf': 36, 'max_bin': 284, 'lambda_l1': 0, 'lambda_l2': 40, 'min_gain_to_split': 1.8228593817250474, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:05:57,164]\u001b[0m Trial 69 finished with value: 0.9992838727814315 and parameters: {'max_depth': 8, 'learning_rate': 0.2701201907507484, 'num_leaves': 255, 'min_data_in_leaf': 83, 'max_bin': 249, 'lambda_l1': 10, 'lambda_l2': 30, 'min_gain_to_split': 1.2764883030333825, 'bagging_fraction': 0.5, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:06:19,220]\u001b[0m Trial 70 finished with value: 0.9993497403831529 and parameters: {'max_depth': 7, 'learning_rate': 0.24290529674720138, 'num_leaves': 58, 'min_data_in_leaf': 128, 'max_bin': 174, 'lambda_l1': 15, 'lambda_l2': 15, 'min_gain_to_split': 0.3185535976408656, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:06:45,270]\u001b[0m Trial 71 finished with value: 0.9996044788104905 and parameters: {'max_depth': 9, 'learning_rate': 0.2603984772271454, 'num_leaves': 73, 'min_data_in_leaf': 25, 'max_bin': 200, 'lambda_l1': 0, 'lambda_l2': 35, 'min_gain_to_split': 0.8900002852948238, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:07:08,758]\u001b[0m Trial 72 finished with value: 0.9995451777687873 and parameters: {'max_depth': 10, 'learning_rate': 0.2610087822417012, 'num_leaves': 95, 'min_data_in_leaf': 24, 'max_bin': 123, 'lambda_l1': 5, 'lambda_l2': 40, 'min_gain_to_split': 0.7716850221057647, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:07:32,331]\u001b[0m Trial 73 finished with value: 0.9995407857856822 and parameters: {'max_depth': 9, 'learning_rate': 0.2866465193320875, 'num_leaves': 70, 'min_data_in_leaf': 51, 'max_bin': 235, 'lambda_l1': 0, 'lambda_l2': 45, 'min_gain_to_split': 1.5164437798880899, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:07:58,022]\u001b[0m Trial 74 finished with value: 0.9996022770751702 and parameters: {'max_depth': 7, 'learning_rate': 0.2747037067610638, 'num_leaves': 43, 'min_data_in_leaf': 64, 'max_bin': 196, 'lambda_l1': 0, 'lambda_l2': 35, 'min_gain_to_split': 0.9153569525795721, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:08:20,246]\u001b[0m Trial 75 finished with value: 0.999430983394932 and parameters: {'max_depth': 10, 'learning_rate': 0.24918143744437873, 'num_leaves': 89, 'min_data_in_leaf': 5, 'max_bin': 196, 'lambda_l1': 5, 'lambda_l2': 30, 'min_gain_to_split': 2.390339637534708, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:08:45,160]\u001b[0m Trial 76 finished with value: 0.9994178089893943 and parameters: {'max_depth': 9, 'learning_rate': 0.2248704237388195, 'num_leaves': 69, 'min_data_in_leaf': 65, 'max_bin': 206, 'lambda_l1': 10, 'lambda_l2': 30, 'min_gain_to_split': 0.9610668974091731, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 43 with value: 0.999740677548943.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:09:18,592]\u001b[0m Trial 77 finished with value: 0.9997582540744461 and parameters: {'max_depth': 8, 'learning_rate': 0.2659848412543876, 'num_leaves': 256, 'min_data_in_leaf': 38, 'max_bin': 218, 'lambda_l1': 0, 'lambda_l2': 40, 'min_gain_to_split': 0.02540093025531065, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:09:39,385]\u001b[0m Trial 78 finished with value: 0.9993277744265463 and parameters: {'max_depth': 8, 'learning_rate': 0.2711319994167853, 'num_leaves': 256, 'min_data_in_leaf': 43, 'max_bin': 189, 'lambda_l1': 15, 'lambda_l2': 20, 'min_gain_to_split': 1.9954701491347886, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:09:56,617]\u001b[0m Trial 79 finished with value: 0.9986364079131219 and parameters: {'max_depth': 9, 'learning_rate': 0.2563793733281495, 'num_leaves': 76, 'min_data_in_leaf': 58, 'max_bin': 218, 'lambda_l1': 0, 'lambda_l2': 40, 'min_gain_to_split': 3.545667314462496, 'bagging_fraction': 0.2, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:10:23,214]\u001b[0m Trial 80 finished with value: 0.999628643456495 and parameters: {'max_depth': 11, 'learning_rate': 0.265252523874844, 'num_leaves': 574, 'min_data_in_leaf': 34, 'max_bin': 204, 'lambda_l1': 0, 'lambda_l2': 45, 'min_gain_to_split': 1.19651927019471, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:10:50,073]\u001b[0m Trial 81 finished with value: 0.9996374278091169 and parameters: {'max_depth': 13, 'learning_rate': 0.26491796548071417, 'num_leaves': 1419, 'min_data_in_leaf': 31, 'max_bin': 211, 'lambda_l1': 0, 'lambda_l2': 45, 'min_gain_to_split': 1.1472686286159604, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:11:15,234]\u001b[0m Trial 82 finished with value: 0.9995583510628219 and parameters: {'max_depth': 13, 'learning_rate': 0.26452125919054387, 'num_leaves': 1373, 'min_data_in_leaf': 33, 'max_bin': 214, 'lambda_l1': 5, 'lambda_l2': 45, 'min_gain_to_split': 1.588183058724013, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:11:45,741]\u001b[0m Trial 83 finished with value: 0.9997033314033367 and parameters: {'max_depth': 13, 'learning_rate': 0.2520992323685065, 'num_leaves': 1619, 'min_data_in_leaf': 18, 'max_bin': 224, 'lambda_l1': 0, 'lambda_l2': 45, 'min_gain_to_split': 0.446339075490216, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:12:18,848]\u001b[0m Trial 84 finished with value: 0.9995671395193536 and parameters: {'max_depth': 13, 'learning_rate': 0.24098937960682054, 'num_leaves': 1438, 'min_data_in_leaf': 18, 'max_bin': 261, 'lambda_l1': 10, 'lambda_l2': 45, 'min_gain_to_split': 0.35547558865499723, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:12:40,517]\u001b[0m Trial 85 finished with value: 0.9991346007956246 and parameters: {'max_depth': 12, 'learning_rate': 0.2513936848826649, 'num_leaves': 852, 'min_data_in_leaf': 48, 'max_bin': 224, 'lambda_l1': 5, 'lambda_l2': 40, 'min_gain_to_split': 1.268661431952242, 'bagging_fraction': 0.30000000000000004, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:13:07,744]\u001b[0m Trial 86 finished with value: 0.9994441589135585 and parameters: {'max_depth': 14, 'learning_rate': 0.23260465103300504, 'num_leaves': 2100, 'min_data_in_leaf': 492, 'max_bin': 241, 'lambda_l1': 0, 'lambda_l2': 55, 'min_gain_to_split': 1.9725559335960199, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:13:32,415]\u001b[0m Trial 87 finished with value: 0.9989413760031169 and parameters: {'max_depth': 12, 'learning_rate': 0.27700056382538524, 'num_leaves': 560, 'min_data_in_leaf': 74, 'max_bin': 207, 'lambda_l1': 65, 'lambda_l2': 55, 'min_gain_to_split': 0.6085931948595914, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:13:52,337]\u001b[0m Trial 88 finished with value: 0.9987128486699435 and parameters: {'max_depth': 11, 'learning_rate': 0.24520147681503726, 'num_leaves': 499, 'min_data_in_leaf': 35, 'max_bin': 224, 'lambda_l1': 5, 'lambda_l2': 50, 'min_gain_to_split': 13.188611804597228, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:14:10,727]\u001b[0m Trial 89 finished with value: 0.9989127169136662 and parameters: {'max_depth': 14, 'learning_rate': 0.22131978764654064, 'num_leaves': 3771, 'min_data_in_leaf': 442, 'max_bin': 180, 'lambda_l1': 10, 'lambda_l2': 45, 'min_gain_to_split': 8.323436959617194, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.30000000000000004}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:14:46,872]\u001b[0m Trial 90 finished with value: 0.9955032037867388 and parameters: {'max_depth': 13, 'learning_rate': 0.02115454294632671, 'num_leaves': 945, 'min_data_in_leaf': 18, 'max_bin': 235, 'lambda_l1': 0, 'lambda_l2': 50, 'min_gain_to_split': 3.2074042887111918, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:15:21,072]\u001b[0m Trial 91 finished with value: 0.999655001100292 and parameters: {'max_depth': 10, 'learning_rate': 0.2649981032456318, 'num_leaves': 389, 'min_data_in_leaf': 17, 'max_bin': 204, 'lambda_l1': 0, 'lambda_l2': 100, 'min_gain_to_split': 0.05912176048350469, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:15:52,854]\u001b[0m Trial 92 finished with value: 0.9996396239937976 and parameters: {'max_depth': 11, 'learning_rate': 0.26701192586789524, 'num_leaves': 1616, 'min_data_in_leaf': 34, 'max_bin': 204, 'lambda_l1': 0, 'lambda_l2': 100, 'min_gain_to_split': 0.2520794592384807, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:16:26,822]\u001b[0m Trial 93 finished with value: 0.9996088730149395 and parameters: {'max_depth': 10, 'learning_rate': 0.28983571231884175, 'num_leaves': 1024, 'min_data_in_leaf': 17, 'max_bin': 215, 'lambda_l1': 5, 'lambda_l2': 100, 'min_gain_to_split': 0.021383720666153738, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:16:55,709]\u001b[0m Trial 94 finished with value: 0.9995847108322804 and parameters: {'max_depth': 11, 'learning_rate': 0.2820670334517253, 'num_leaves': 1553, 'min_data_in_leaf': 48, 'max_bin': 190, 'lambda_l1': 5, 'lambda_l2': 95, 'min_gain_to_split': 0.37470294226867895, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:17:27,423]\u001b[0m Trial 95 finished with value: 0.999510042141163 and parameters: {'max_depth': 10, 'learning_rate': 0.2038106385157768, 'num_leaves': 372, 'min_data_in_leaf': 218, 'max_bin': 224, 'lambda_l1': 0, 'lambda_l2': 95, 'min_gain_to_split': 0.2649757487111414, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:17:54,888]\u001b[0m Trial 96 finished with value: 0.9994178036312376 and parameters: {'max_depth': 12, 'learning_rate': 0.2681434425441717, 'num_leaves': 1810, 'min_data_in_leaf': 92, 'max_bin': 243, 'lambda_l1': 10, 'lambda_l2': 95, 'min_gain_to_split': 0.7203497196831435, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:18:20,784]\u001b[0m Trial 97 finished with value: 0.9989260018764717 and parameters: {'max_depth': 14, 'learning_rate': 0.25436073476762616, 'num_leaves': 3136, 'min_data_in_leaf': 31, 'max_bin': 184, 'lambda_l1': 85, 'lambda_l2': 80, 'min_gain_to_split': 0.0195468424433225, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:18:42,184]\u001b[0m Trial 98 finished with value: 0.9902735900903838 and parameters: {'max_depth': 2, 'learning_rate': 0.2751011120975684, 'num_leaves': 4, 'min_data_in_leaf': 6, 'max_bin': 212, 'lambda_l1': 0, 'lambda_l2': 90, 'min_gain_to_split': 0.6294848547301051, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n","\u001b[32m[I 2022-07-31 01:19:07,091]\u001b[0m Trial 99 finished with value: 0.9994199970633535 and parameters: {'max_depth': 9, 'learning_rate': 0.25813175614135114, 'num_leaves': 202, 'min_data_in_leaf': 15, 'max_bin': 231, 'lambda_l1': 5, 'lambda_l2': 100, 'min_gain_to_split': 1.6336804228561772, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 77 with value: 0.9997582540744461.\u001b[0m\n"]}]},{"cell_type":"code","source":["print(f\"\\tBest value (Accuracy): {study.best_value:.5f}\")\n","print(f\"\\tBest params:\")\n","\n","for key, value in study.best_params.items():\n","    print(f\"\\t\\t{key}: {value}\")"],"metadata":{"id":"3hbvJEcuFvrB"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["param_lgb = {'max_depth': 13,\n","\t\t'learning_rate': 0.22690320686746146,\n","\t\t'num_leaves': 501,\n","\t\t'min_data_in_leaf': 168,\n","\t\t'max_bin': 61,\n","\t\t'lambda_l1': 0,\n","\t\t'lambda_l2': 0,\n","\t\t'min_gain_to_split': 1.2780588498979437,\n","\t\t'bagging_fraction': 0.9,\n","\t\t'bagging_freq': 1,\n","\t\t'feature_fraction': 0.2    \n","}\n","\n","lgb_opt = lgb.LGBMClassifier(**param_lgb)\n","run_model(lgb_opt, X_train_smt, y_train_smt, X_test, y_test)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":956},"id":"37p80nFY-olc","outputId":"20fe5997-af2c-4f49-f0e6-816988c5de4b","executionInfo":{"status":"ok","timestamp":1659233417761,"user_tz":-420,"elapsed":7855,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}}},"execution_count":100,"outputs":[{"output_type":"stream","name":"stdout","text":["Accuracy = 0.9992977774656788\n","ROC Area under Curve = 0.93046444160077\n","F1 Score  = 0.8130841121495327\n","Cohen's Kappa = 0.8127334482777326\n","Time taken = 6.331333160400391\n","              precision    recall  f1-score   support\n","\n","           0    0.99975   0.99954   0.99965     56861\n","           1    0.76991   0.86139   0.81308       101\n","\n","    accuracy                        0.99930     56962\n","   macro avg    0.88483   0.93046   0.90637     56962\n","weighted avg    0.99935   0.99930   0.99932     56962\n","\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}},{"output_type":"execute_result","data":{"text/plain":["(LGBMClassifier(bagging_fraction=0.9, bagging_freq=1, feature_fraction=0.2,\n","                lambda_l1=0, lambda_l2=0, learning_rate=0.22690320686746146,\n","                max_bin=61, max_depth=13, min_data_in_leaf=168,\n","                min_gain_to_split=1.2780588498979437, num_leaves=501),\n"," 0.9992977774656788,\n"," 0.93046444160077,\n"," 0.8130841121495327,\n"," 0.8127334482777326,\n"," 6.331333160400391)"]},"metadata":{},"execution_count":100},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":["import plotly.offline as py\n","py.init_notebook_mode(connected=True)\n","import plotly.graph_objs as go\n","import plotly.tools as tls\n","import plotly.figure_factory as ff\n","import plotly.io as pio\n","pio.renderers.default='colab'\n","\n","def model_performance(model, X_test, y_test):\n","    y_pred = model.predict(X_test)\n","    y_score = model.predict_proba(X_test)[:,1]\n","\n","    #Conf matrix\n","    conf_matrix = confusion_matrix(y_test, y_pred)\n","    trace1 = go.Heatmap(z = conf_matrix  ,x = [\"0 (pred)\",\"1 (pred)\"],\n","                        y = [\"0 (true)\",\"1 (true)\"],xgap = 2, ygap = 2,\n","                        text = conf_matrix,\n","                        colorscale = 'Viridis', showscale  = False)\n","\n","    #Show metrics\n","    tp = conf_matrix[1,1]\n","    fn = conf_matrix[1,0]\n","    fp = conf_matrix[0,1]\n","    tn = conf_matrix[0,0]\n","    Accuracy  =  ((tp+tn)/(tp+tn+fp+fn))\n","    Precision =  (tp/(tp+fp))\n","    Recall    =  (tp/(tp+fn))\n","    F1_score  =  (2*(((tp/(tp+fp))*(tp/(tp+fn)))/((tp/(tp+fp))+(tp/(tp+fn)))))\n","\n","    show_metrics = pd.DataFrame(data=[[Accuracy , Precision, Recall, F1_score]])\n","    show_metrics = show_metrics.T\n","\n","    colors = ['gold', 'lightgreen', 'lightcoral', 'lightskyblue']\n","    trace2 = go.Bar(x = (show_metrics[0].values), \n","                   y = ['Accuracy', 'Precision', 'Recall', 'F1_score'], text = np.round_(show_metrics[0].values,4),\n","                    textposition = 'auto',\n","                   orientation = 'h', opacity = 0.8,marker=dict(\n","            color=colors,\n","            line=dict(color='#000000',width=1.5)))\n","    \n","    #Roc curve\n","    model_roc_auc = round(roc_auc_score(y_test, y_score) , 3)\n","    fpr, tpr, t = roc_curve(y_test, y_score)\n","    trace3 = go.Scatter(x = fpr,y = tpr,\n","                        name = \"Roc : \" + str(model_roc_auc),\n","                        line = dict(color = ('rgb(22, 96, 167)'),width = 2), fill='tozeroy')\n","    trace4 = go.Scatter(x = [0,1],y = [0,1],\n","                        line = dict(color = ('black'),width = 1.5,\n","                        dash = 'dot'))\n","    \n","    # Precision-recall curve\n","    precision, recall, thresholds = precision_recall_curve(y_test, y_score)\n","    trace5 = go.Scatter(x = recall, y = precision,\n","                        name = \"Precision\" + str(precision),\n","                        line = dict(color = ('lightcoral'),width = 2), fill='tozeroy')\n","    \n","    #Feature importance\n","    coefficients  = pd.DataFrame(model.feature_importances_)\n","    column_data   = pd.DataFrame(list(df))\n","    coef_sumry    = (pd.merge(coefficients,column_data,left_index= True,\n","                              right_index= True, how = \"left\"))\n","    coef_sumry.columns = [\"coefficients\",\"features\"]\n","    coef_sumry    = coef_sumry.sort_values(by = \"coefficients\",ascending = False)\n","    coef_sumry = coef_sumry[coef_sumry[\"coefficients\"] !=0]\n","    trace6 = go.Bar(x = coef_sumry[\"features\"],y = coef_sumry[\"coefficients\"],\n","                    name = \"coefficients\",\n","                    marker = dict(color = coef_sumry[\"coefficients\"],\n","                                  colorscale = \"Viridis\",\n","                                  line = dict(width = .6,color = \"black\")))\n","    \n","    #Cumulative gain\n","    pos = pd.get_dummies(y_test).values\n","    pos = pos[:,1] \n","    npos = np.sum(pos)\n","    index = np.argsort(y_score) \n","    index = index[::-1] \n","    sort_pos = pos[index]\n","    #cumulative sum\n","    cpos = np.cumsum(sort_pos) \n","    #recall\n","    recall = cpos/npos \n","    #size obs test\n","    n = y_test.shape[0] \n","    size = np.arange(start=1,stop=369,step=1) \n","    #proportion\n","    size = size / n \n","    #plots\n","    model = model\n","    trace7 = go.Scatter(x = size,y = recall,\n","                        name = \"Lift curve\",\n","                        line = dict(color = ('gold'),width = 2), fill='tozeroy') \n","    \n","    #Subplots\n","    fig = tls.make_subplots(rows=4, cols=2, print_grid=False, \n","                          specs=[[{}, {}], \n","                                 [{}, {}],\n","                                 [{'colspan': 2}, None],\n","                                 [{'colspan': 2}, None]],\n","                          subplot_titles=('Confusion Matrix',\n","                                        'Metrics',\n","                                        'ROC curve'+\" \"+ '('+ str(model_roc_auc)+')',\n","                                        'Precision - Recall curve',\n","                                        'Cumulative gains curve',\n","                                        'Feature importance',\n","                                        ))\n","    \n","    fig.append_trace(trace1,1,1)\n","    fig.append_trace(trace2,1,2)\n","    fig.append_trace(trace3,2,1)\n","    fig.append_trace(trace4,2,1)\n","    fig.append_trace(trace5,2,2)\n","    fig.append_trace(trace6,4,1)\n","    fig.append_trace(trace7,3,1)\n","    \n","    fig['layout'].update(showlegend = False, title = '<b>Model performance report</b><br>' + 'LightGBM after tuning',\n","                        autosize = False, height = 1500,width = 830,\n","                        plot_bgcolor = 'rgba(240,240,240, 0.95)',\n","                        paper_bgcolor = 'rgba(240,240,240, 0.95)',\n","                        margin = dict(b = 195))\n","    fig[\"layout\"][\"xaxis2\"].update((dict(range=[0, 1])))\n","    fig[\"layout\"][\"xaxis3\"].update(dict(title = \"false positive rate\"))\n","    fig[\"layout\"][\"yaxis3\"].update(dict(title = \"true positive rate\"))\n","    fig[\"layout\"][\"xaxis4\"].update(dict(title = \"recall\"), range = [0,1.05])\n","    fig[\"layout\"][\"yaxis4\"].update(dict(title = \"precision\"), range = [0,1.05])\n","    fig[\"layout\"][\"xaxis5\"].update(dict(title = \"Percentage contacted\"))\n","    fig[\"layout\"][\"yaxis5\"].update(dict(title = \"Percentage positive targeted\"))\n","    fig.layout.titlefont.size = 14\n","    \n","    fig.show(\"colab\")\n","    # pio.write_image(fig, \"op.png\")"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":17},"id":"SsJ1havr5lMi","executionInfo":{"status":"ok","timestamp":1659233919435,"user_tz":-420,"elapsed":304,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}},"outputId":"982ebf75-d6c4-4a11-c927-97c069d72f3a"},"execution_count":110,"outputs":[{"output_type":"display_data","data":{"text/html":["        <script type=\"text/javascript\">\n","        window.PlotlyConfig = {MathJaxConfig: 'local'};\n","        if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}\n","        if (typeof require !== 'undefined') {\n","        require.undef(\"plotly\");\n","        requirejs.config({\n","            paths: {\n","                'plotly': ['https://cdn.plot.ly/plotly-2.8.3.min']\n","            }\n","        });\n","        require(['plotly'], function(Plotly) {\n","            window._Plotly = Plotly;\n","        });\n","        }\n","        </script>\n","        "]},"metadata":{}}]},{"cell_type":"code","source":["model_performance(lgb_opt, X_test, y_test)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000,"output_embedded_package_id":"1DMGPv_cPAIHnDmCfSvjAhIO6yqZ-bYKv"},"id":"BX78gKhn5o4y","executionInfo":{"status":"ok","timestamp":1659233925115,"user_tz":-420,"elapsed":2379,"user":{"displayName":"Nguyen Son","userId":"12638499436909749549"}},"outputId":"33f636c1-a0ff-4d24-c2a8-4a32e352bb39"},"execution_count":111,"outputs":[{"output_type":"display_data","data":{"text/plain":"Output hidden; open in https://colab.research.google.com to view."},"metadata":{}}]},{"cell_type":"markdown","source":["**Observation:**\n","* The model has improved significantly with the `f1_score` index increased from 65.3% to 81.3%\n","* With the Credit Card Fraud Detection problem, we should focus on reducing the number of Type I (FP) errors that the model predicts. Customers fall into the category I error about the fact that they committed fraud, but the model predicts that these customers do not commit fraud. This leads to a great loss for credit institutions when we have wrongly predicted these customers.\n","* Therefore, in addition to choosing a good model, we also need to choose the models with the lowest error types of type I.\n","\n","=> The `LightGBM` model will be the chosen model to solve this problem!"],"metadata":{"id":"l-tGUNzlFrrQ"}}]}