{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "Let's have a look at how to implement a decision tree in Python. Again, we need to first import the required packages and load the data" ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:42:11.447517Z", "iopub.status.busy": "2024-05-31T21:42:11.447229Z", "iopub.status.idle": "2024-05-31T21:42:14.385646Z", "shell.execute_reply": "2024-05-31T21:42:14.384714Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Dataset already downloaded!\n" ] } ], "source": [ "import pandas as pd\n", "import numpy as np\n", "import matplotlib.pyplot as plt\n", "import seaborn as sns\n", "from sklearn.preprocessing import StandardScaler, MinMaxScaler\n", "from sklearn.tree import DecisionTreeClassifier, plot_tree\n", "from sklearn.ensemble import RandomForestClassifier\n", "from xgboost import XGBClassifier\n", "from sklearn.model_selection import train_test_split\n", "from sklearn.metrics import confusion_matrix, accuracy_score, roc_auc_score, recall_score, precision_score, roc_curve\n", "from sklearn.inspection import permutation_importance\n", "pd.set_option('display.max_columns', 50) # Display up to 50 columns\n", "from io import BytesIO\n", "from urllib.request import urlopen\n", "from zipfile import ZipFile\n", "import os.path\n", "\n", "# Check if the file exists\n", "if not os.path.isfile('data/card_transdata.csv'):\n", "\n", " print('Downloading dataset...')\n", "\n", " # Define the dataset to be downloaded\n", " zipurl = 'https://www.kaggle.com/api/v1/datasets/download/dhanushnarayananr/credit-card-fraud'\n", "\n", " # Download and unzip the dataset in the data folder\n", " with urlopen(zipurl) as zipresp:\n", " with ZipFile(BytesIO(zipresp.read())) as zfile:\n", " zfile.extractall('data')\n", "\n", " print('DONE!')\n", "\n", "else:\n", "\n", " print('Dataset already downloaded!')\n", "\n", "# Load the data\n", "df = pd.read_csv('data/card_transdata.csv')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This is the **dataset of credit card transactions** from [Kaggle.com](https://www.kaggle.com/datasets/dhanushnarayananr/credit-card-fraud/data) which we have used before. Recall that the target variable $y$ is `fraud`, which indicates whether the transaction is fraudulent or not. The other variables are the features $x$ of the transactions." ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:42:14.389689Z", "iopub.status.busy": "2024-05-31T21:42:14.389392Z", "iopub.status.idle": "2024-05-31T21:42:14.413731Z", "shell.execute_reply": "2024-05-31T21:42:14.413088Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
distance_from_homedistance_from_last_transactionratio_to_median_purchase_pricerepeat_retailerused_chipused_pin_numberonline_orderfraud
057.8778570.3111401.9459401.01.00.00.00.0
110.8299430.1755921.2942191.00.00.00.00.0
25.0910790.8051530.4277151.00.00.01.00.0
32.2475645.6000440.3626631.01.00.01.00.0
444.1909360.5664862.2227671.01.00.01.00.0
55.58640813.2610730.0647681.00.00.00.00.0
63.7240190.9568380.2784651.00.00.01.00.0
74.8482470.3207351.2730501.00.01.00.00.0
80.8766322.5036091.5169990.00.00.00.00.0
98.8390472.9705122.3616831.00.00.01.00.0
1014.2635300.1587581.1361021.01.00.01.00.0
1113.5923680.2405401.3703301.01.00.01.00.0
12765.2825590.3715620.5512451.01.00.00.00.0
132.13195656.3724016.3586671.00.00.01.01.0
1413.9559720.2715222.7989011.00.00.01.00.0
15179.6651480.1209200.5356401.01.01.01.00.0
16114.5197890.7070030.5169901.00.00.00.00.0
173.5896496.2474581.8464511.00.00.00.00.0
1811.08515234.6613512.5307581.00.00.01.00.0
196.1946711.1420140.3072171.00.00.00.00.0
\n", "
" ], "text/plain": [ " distance_from_home distance_from_last_transaction \\\n", "0 57.877857 0.311140 \n", "1 10.829943 0.175592 \n", "2 5.091079 0.805153 \n", "3 2.247564 5.600044 \n", "4 44.190936 0.566486 \n", "5 5.586408 13.261073 \n", "6 3.724019 0.956838 \n", "7 4.848247 0.320735 \n", "8 0.876632 2.503609 \n", "9 8.839047 2.970512 \n", "10 14.263530 0.158758 \n", "11 13.592368 0.240540 \n", "12 765.282559 0.371562 \n", "13 2.131956 56.372401 \n", "14 13.955972 0.271522 \n", "15 179.665148 0.120920 \n", "16 114.519789 0.707003 \n", "17 3.589649 6.247458 \n", "18 11.085152 34.661351 \n", "19 6.194671 1.142014 \n", "\n", " ratio_to_median_purchase_price repeat_retailer used_chip \\\n", "0 1.945940 1.0 1.0 \n", "1 1.294219 1.0 0.0 \n", "2 0.427715 1.0 0.0 \n", "3 0.362663 1.0 1.0 \n", "4 2.222767 1.0 1.0 \n", "5 0.064768 1.0 0.0 \n", "6 0.278465 1.0 0.0 \n", "7 1.273050 1.0 0.0 \n", "8 1.516999 0.0 0.0 \n", "9 2.361683 1.0 0.0 \n", "10 1.136102 1.0 1.0 \n", "11 1.370330 1.0 1.0 \n", "12 0.551245 1.0 1.0 \n", "13 6.358667 1.0 0.0 \n", "14 2.798901 1.0 0.0 \n", "15 0.535640 1.0 1.0 \n", "16 0.516990 1.0 0.0 \n", "17 1.846451 1.0 0.0 \n", "18 2.530758 1.0 0.0 \n", "19 0.307217 1.0 0.0 \n", "\n", " used_pin_number online_order fraud \n", "0 0.0 0.0 0.0 \n", "1 0.0 0.0 0.0 \n", "2 0.0 1.0 0.0 \n", "3 0.0 1.0 0.0 \n", "4 0.0 1.0 0.0 \n", "5 0.0 0.0 0.0 \n", "6 0.0 1.0 0.0 \n", "7 1.0 0.0 0.0 \n", "8 0.0 0.0 0.0 \n", "9 0.0 1.0 0.0 \n", "10 0.0 1.0 0.0 \n", "11 0.0 1.0 0.0 \n", "12 0.0 0.0 0.0 \n", "13 0.0 1.0 1.0 \n", "14 0.0 1.0 0.0 \n", "15 1.0 1.0 0.0 \n", "16 0.0 0.0 0.0 \n", "17 0.0 0.0 0.0 \n", "18 0.0 1.0 0.0 \n", "19 0.0 0.0 0.0 " ] }, "execution_count": 2, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.head(20)" ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:42:14.416833Z", "iopub.status.busy": "2024-05-31T21:42:14.416577Z", "iopub.status.idle": "2024-05-31T21:42:14.681122Z", "shell.execute_reply": "2024-05-31T21:42:14.680506Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
distance_from_homedistance_from_last_transactionratio_to_median_purchase_pricerepeat_retailerused_chipused_pin_numberonline_orderfraud
count1000000.0000001000000.0000001000000.0000001000000.0000001000000.0000001000000.0000001000000.0000001000000.000000
mean26.6287925.0365191.8241820.8815360.3503990.1006080.6505520.087403
std65.39078425.8430932.7995890.3231570.4770950.3008090.4767960.282425
min0.0048740.0001180.0043990.0000000.0000000.0000000.0000000.000000
25%3.8780080.2966710.4756731.0000000.0000000.0000000.0000000.000000
50%9.9677600.9986500.9977171.0000000.0000000.0000001.0000000.000000
75%25.7439853.3557482.0963701.0000001.0000000.0000001.0000000.000000
max10632.72367211851.104565267.8029421.0000001.0000001.0000001.0000001.000000
\n", "
" ], "text/plain": [ " distance_from_home distance_from_last_transaction \\\n", "count 1000000.000000 1000000.000000 \n", "mean 26.628792 5.036519 \n", "std 65.390784 25.843093 \n", "min 0.004874 0.000118 \n", "25% 3.878008 0.296671 \n", "50% 9.967760 0.998650 \n", "75% 25.743985 3.355748 \n", "max 10632.723672 11851.104565 \n", "\n", " ratio_to_median_purchase_price repeat_retailer used_chip \\\n", "count 1000000.000000 1000000.000000 1000000.000000 \n", "mean 1.824182 0.881536 0.350399 \n", "std 2.799589 0.323157 0.477095 \n", "min 0.004399 0.000000 0.000000 \n", "25% 0.475673 1.000000 0.000000 \n", "50% 0.997717 1.000000 0.000000 \n", "75% 2.096370 1.000000 1.000000 \n", "max 267.802942 1.000000 1.000000 \n", "\n", " used_pin_number online_order fraud \n", "count 1000000.000000 1000000.000000 1000000.000000 \n", "mean 0.100608 0.650552 0.087403 \n", "std 0.300809 0.476796 0.282425 \n", "min 0.000000 0.000000 0.000000 \n", "25% 0.000000 0.000000 0.000000 \n", "50% 0.000000 1.000000 0.000000 \n", "75% 0.000000 1.000000 0.000000 \n", "max 1.000000 1.000000 1.000000 " ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.describe()" ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:42:14.684192Z", "iopub.status.busy": "2024-05-31T21:42:14.683945Z", "iopub.status.idle": "2024-05-31T21:42:14.706136Z", "shell.execute_reply": "2024-05-31T21:42:14.705541Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "RangeIndex: 1000000 entries, 0 to 999999\n", "Data columns (total 8 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 distance_from_home 1000000 non-null float64\n", " 1 distance_from_last_transaction 1000000 non-null float64\n", " 2 ratio_to_median_purchase_price 1000000 non-null float64\n", " 3 repeat_retailer 1000000 non-null float64\n", " 4 used_chip 1000000 non-null float64\n", " 5 used_pin_number 1000000 non-null float64\n", " 6 online_order 1000000 non-null float64\n", " 7 fraud 1000000 non-null float64\n", "dtypes: float64(8)\n", "memory usage: 61.0 MB\n" ] } ], "source": [ "df.info()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Data Preprocessing\n", "\n", "Since we have already explored the dataset in the previous notebook, we can skip that part and move directly to the data preprocessing. \n", "\n", "We will again split the data into training and test sets using the `train_test_split` function" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:42:14.709129Z", "iopub.status.busy": "2024-05-31T21:42:14.708889Z", "iopub.status.idle": "2024-05-31T21:42:15.140280Z", "shell.execute_reply": "2024-05-31T21:42:15.139363Z" } }, "outputs": [], "source": [ "X = df.drop('fraud', axis=1) # All variables except `fraud`\n", "y = df['fraud'] # Only our fraud variables\n", "X_train, X_test, y_train, y_test = train_test_split(X, y, stratify=y, test_size = 0.3, random_state = 42)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Then we can do the feature scaling to ensure our non-binary variables have mean zero and variance 1" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:42:15.144210Z", "iopub.status.busy": "2024-05-31T21:42:15.143906Z", "iopub.status.idle": "2024-05-31T21:42:15.212285Z", "shell.execute_reply": "2024-05-31T21:42:15.211306Z" } }, "outputs": [], "source": [ "def scale_features(scaler, df, col_names, only_transform=False):\n", "\n", " # Extract the features we want to scale\n", " features = df[col_names] \n", "\n", " # Fit the scaler to the features and transform them\n", " if only_transform:\n", " features = scaler.transform(features.values)\n", " else:\n", " features = scaler.fit_transform(features.values)\n", "\n", " # Replace the original features with the scaled features\n", " df[col_names] = features\n", "\n", "col_names = ['distance_from_home', 'distance_from_last_transaction', 'ratio_to_median_purchase_price'] \n", "scaler = StandardScaler() \n", "scale_features(scaler, X_train, col_names)\n", "scale_features(scaler, X_test, col_names, only_transform=True)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Implementing a Decision Tree Classifier\n", "\n", "We can now implement a decision tree model using the `DecisionTreeClassifier` class from the `sklearn.tree` module. Fitting the model to the data is almost the same as when we used logistic regression" ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:42:15.216478Z", "iopub.status.busy": "2024-05-31T21:42:15.216062Z", "iopub.status.idle": "2024-05-31T21:42:16.810704Z", "shell.execute_reply": "2024-05-31T21:42:16.809876Z" } }, "outputs": [], "source": [ "clf_dt = DecisionTreeClassifier(random_state=0).fit(X_train, y_train)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can visualize the tree using the `plot_tree` function from the `sklearn.tree` module" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:42:16.814207Z", "iopub.status.busy": "2024-05-31T21:42:16.813956Z", "iopub.status.idle": "2024-05-31T21:42:18.615317Z", "shell.execute_reply": "2024-05-31T21:42:18.614624Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_tree(clf_dt, filled=True, feature_names = X_train.columns.to_list())\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The tree is quite large and it's difficult to see details. Let's only look at the first level of the tree" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:42:18.618482Z", "iopub.status.busy": "2024-05-31T21:42:18.618226Z", "iopub.status.idle": "2024-05-31T21:42:18.837833Z", "shell.execute_reply": "2024-05-31T21:42:18.837226Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_tree(clf_dt, max_depth=1, filled=True, feature_names = X_train.columns.to_list(), fontsize=10)\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Recall from the data exploration that `ratio_to_median_purchase_price` was highly correlated with fraud. The decision tree model seems to have picked up on this as well since the first split is based on this variable. Also, note that the order in which the variables are split can differ between different branches of the tree.\n", "\n", "We can also make predictions using the model and evaluate its performance using the same functions as before" ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:42:18.841487Z", "iopub.status.busy": "2024-05-31T21:42:18.841201Z", "iopub.status.idle": "2024-05-31T21:42:19.336618Z", "shell.execute_reply": "2024-05-31T21:42:19.335912Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Accuracy: 0.9999833333333333\n", "Precision: 0.9999237223493517\n", "Recall: 0.999885587887571\n", "ROC AUC: 0.999939141362689\n" ] } ], "source": [ "y_pred_dt = clf_dt.predict(X_test)\n", "y_proba_dt = clf_dt.predict_proba(X_test)\n", "\n", "print(f\"Accuracy: {accuracy_score(y_test, y_pred_dt)}\")\n", "print(f\"Precision: {precision_score(y_test, y_pred_dt)}\")\n", "print(f\"Recall: {recall_score(y_test, y_pred_dt)}\")\n", "print(f\"ROC AUC: {roc_auc_score(y_test, y_proba_dt[:, 1])}\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The decision tree performs substantially better than the logistic regression. The ROC AUC score is much closer to the maximum value of 1 and we have an almost perfect classifier" ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:42:19.340610Z", "iopub.status.busy": "2024-05-31T21:42:19.340303Z", "iopub.status.idle": "2024-05-31T21:42:19.547348Z", "shell.execute_reply": "2024-05-31T21:42:19.546207Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Compute the ROC curve\n", "fpr, tpr, thresholds = roc_curve(y_test, y_proba_dt[:, 1])\n", "\n", "# Plot the ROC curve\n", "plt.plot(fpr, tpr)\n", "plt.plot([0, 1], [0, 1], linestyle='--', color='grey')\n", "plt.xlabel('False Positive Rate (FPR)')\n", "plt.ylabel('True Positive Rate (TPR)')\n", "plt.title('ROC Curve')\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's also check the confusion matrix to see where we still make mistakes" ] }, { "cell_type": "code", "execution_count": 12, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:42:19.550818Z", "iopub.status.busy": "2024-05-31T21:42:19.550570Z", "iopub.status.idle": "2024-05-31T21:42:19.929916Z", "shell.execute_reply": "2024-05-31T21:42:19.929072Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "conf_mat = confusion_matrix(y_test, y_pred_dt, labels=[1, 0]).transpose() # Transpose the sklearn confusion matrix to match the convention in the lecture\n", "sns.heatmap(conf_mat, annot=True, cmap='Blues', fmt='g', xticklabels=['Fraud', 'No Fraud'], yticklabels=['Fraud', 'No Fraud'])\n", "plt.xlabel(\"Actual\")\n", "plt.ylabel(\"Predicted\")\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "There are only 3 false negatives, i.e., fraudulent transactions that we did not detect. There are also 2 false positives, i.e., \"false alarms\", where non-fraudulent transactions were classified as fraudulent. The decision tree classifier is almost perfect which is a bit suspicious. We might have been lucky in the sense that the training and test sets were split in a way that the model performs very well. We should not expect this to be the case in general. It might be better to use cross-validation to get a more reliable estimate of the model's performance.\n", "\n", "\n", "### Implementing a Random Forest Classifier\n", "\n", "We can also implement a random forest model using the `RandomForestClassifier` class from the `sklearn.ensemble` module. Fitting the model to the data is almost the same as when we used logistic regression and decision trees" ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:42:19.934623Z", "iopub.status.busy": "2024-05-31T21:42:19.934275Z", "iopub.status.idle": "2024-05-31T21:43:30.233105Z", "shell.execute_reply": "2024-05-31T21:43:30.232143Z" } }, "outputs": [], "source": [ "clf_rf = RandomForestClassifier(random_state = 0).fit(X_train, y_train)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Note that it takes a bit longer to train the Random Forest since we have to train many trees (the default setting is 100). We can also make predictions using the model and evaluate its performance using the same functions as before" ] }, { "cell_type": "code", "execution_count": 14, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:43:30.236995Z", "iopub.status.busy": "2024-05-31T21:43:30.236695Z", "iopub.status.idle": "2024-05-31T21:43:33.920027Z", "shell.execute_reply": "2024-05-31T21:43:33.919371Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Accuracy: 0.9999833333333333\n", "Precision: 1.0\n", "Recall: 0.9998093131459517\n", "ROC AUC: 0.9999999993035008\n" ] } ], "source": [ "y_pred_rf = clf_rf.predict(X_test)\n", "y_proba_rf = clf_rf.predict_proba(X_test)\n", "\n", "print(f\"Accuracy: {accuracy_score(y_test, y_pred_rf)}\")\n", "print(f\"Precision: {precision_score(y_test, y_pred_rf)}\")\n", "print(f\"Recall: {recall_score(y_test, y_pred_rf)}\")\n", "print(f\"ROC AUC: {roc_auc_score(y_test, y_proba_rf[:, 1])}\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "As expected, the Random Forest performs better than the Decision Tree in the metrics we have used. Now, let's also check the confusion matrix to see where we still make mistakes" ] }, { "cell_type": "code", "execution_count": 15, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:43:33.923169Z", "iopub.status.busy": "2024-05-31T21:43:33.922912Z", "iopub.status.idle": "2024-05-31T21:43:34.283529Z", "shell.execute_reply": "2024-05-31T21:43:34.282757Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "conf_mat = confusion_matrix(y_test, y_pred_rf, labels=[1, 0]).transpose() # Transpose the sklearn confusion matrix to match the convention in the lecture\n", "sns.heatmap(conf_mat, annot=True, cmap='Blues', fmt='g', xticklabels=['Fraud', 'No Fraud'], yticklabels=['Fraud', 'No Fraud'])\n", "plt.xlabel(\"Actual\")\n", "plt.ylabel(\"Predicted\")\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "There are still some false negatives, but the number of false positives has decreased compared to the Decision Tree model. \n", "\n", "\n", "### Implementing a XGBoost Classifier\n", "\n", "Let's also have a look at the XGBoost classifier. We can implement the model using the `XGBClassifier` class from the `xgboost` package. Fitting the model to the data is almost the same as when we used logistic regression, decision trees, and random forests, even though it is not part of the `sklearn` package. This is because the `xgboost` package is designed to work well with the `sklearn` package. Let's fit the model to the data" ] }, { "cell_type": "code", "execution_count": 16, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:43:34.286855Z", "iopub.status.busy": "2024-05-31T21:43:34.286641Z", "iopub.status.idle": "2024-05-31T21:43:36.494100Z", "shell.execute_reply": "2024-05-31T21:43:36.493306Z" } }, "outputs": [], "source": [ "clf_xgb = XGBClassifier(random_state = 0).fit(X_train, y_train)" ] }, { "cell_type": "code", "execution_count": 17, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:43:36.498147Z", "iopub.status.busy": "2024-05-31T21:43:36.497858Z", "iopub.status.idle": "2024-05-31T21:43:37.082068Z", "shell.execute_reply": "2024-05-31T21:43:37.081474Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Accuracy: 0.9983366666666667\n", "Precision: 0.9893835616438356\n", "Recall: 0.9916097784218756\n", "ROC AUC: 0.999973496046352\n" ] } ], "source": [ "y_pred_xgb = clf_xgb.predict(X_test)\n", "y_proba_xgb = clf_xgb.predict_proba(X_test)\n", "\n", "print(f\"Accuracy: {accuracy_score(y_test, y_pred_xgb)}\")\n", "print(f\"Precision: {precision_score(y_test, y_pred_xgb)}\")\n", "print(f\"Recall: {recall_score(y_test, y_pred_xgb)}\")\n", "print(f\"ROC AUC: {roc_auc_score(y_test, y_proba_xgb[:, 1])}\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's also check the confusion matrix to see where we still make mistakes" ] }, { "cell_type": "code", "execution_count": 18, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:43:37.085261Z", "iopub.status.busy": "2024-05-31T21:43:37.085000Z", "iopub.status.idle": "2024-05-31T21:43:37.407143Z", "shell.execute_reply": "2024-05-31T21:43:37.406289Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "conf_mat = confusion_matrix(y_test, y_pred_xgb, labels=[1, 0]).transpose() # Transpose the sklearn confusion matrix to match the convention in the lecture\n", "sns.heatmap(conf_mat, annot=True, cmap='Blues', fmt='g', xticklabels=['Fraud', 'No Fraud'], yticklabels=['Fraud', 'No Fraud'])\n", "plt.xlabel(\"Actual\")\n", "plt.ylabel(\"Predicted\")\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The XGBoost model seems to perform a bit worse than the Random Forest model. There are more false negatives and false positives. However, the model is still very good at detecting fraudulent transactions and has a high ROC AUC score. Adjusting the hyperparameters of the model might improve its performance.\n", "\n", "\n", "### Feature Importance\n", "\n", "We can also look at the feature importance of each model. The feature importance is a measure of how much each feature contributes to the model's predictions. Let's start with the Decision Tree model" ] }, { "cell_type": "code", "execution_count": 19, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:43:37.410221Z", "iopub.status.busy": "2024-05-31T21:43:37.409982Z", "iopub.status.idle": "2024-05-31T21:43:37.558319Z", "shell.execute_reply": "2024-05-31T21:43:37.557737Z" } }, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAwIAAAHFCAYAAABBzk6dAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAABt9klEQVR4nO3deVxO6f8/8Nfdvq9aSAsSpUKyxFDWKNuMsZMwdpNs4UNUtmxjm2EMo2KMZcYyGPuSLYwia2NJYUzGXgpR9/n94df5uttLaTmv5+NxHuNc5zrXeV/n3MZ53+c61y0TBEEAERERERFJilJZB0BERERERJ8fEwEiIiIiIgliIkBEREREJEFMBIiIiIiIJIiJABERERGRBDERICIiIiKSICYCREREREQSxESAiIiIiEiCmAgQEREREUkQEwEiogooPDwcMpks12XSpEmlcswbN24gKCgIiYmJpdL+p0hMTIRMJsPixYvLOpRii4qKQlBQEF6+fFnWoZSJjz/DysrKMDQ0RP369TFixAicO3eu1I/v4eEBDw+PIu0TGRkJmUyGyMjIUokpv2MWZiEqiEpZB0BERMUXFhaGunXrKpRVq1atVI5148YNBAcHw8PDAzY2NqVyDCmLiopCcHAwfH19YWBgUNbhlImvv/4aEydOhCAISElJwbVr17Bhwwb89NNP8PPzw/Lly0vt2KtWrSryPi4uLjh79iwcHBxKIaL8j/mxL7/8ErVq1arQiTCVDSYCREQVmKOjI1xdXcs6jE/y/v17yGQyqKhI85+kN2/eQENDo6zDKBfMzMzQrFkzcd3T0xP+/v4YPnw4VqxYgbp162LUqFGlcuzi3Mzr6ekpxPs55HZMdXV1GBgY5BuLIAh4+/YtNDU1SztEqkA4NIiIqBLbunUr3NzcoK2tDR0dHXh6euLSpUsKdaKjo9GnTx/Y2NhAU1MTNjY26Nu3L+7duyfWCQ8PR8+ePQEArVu3FocehIeHAwBsbGzg6+ub4/jZh1tkDWvYuHEjJk6cCAsLC6irq+POnTsAgCNHjqBt27bQ09ODlpYWWrRogaNHjxar71nDp44dO4Zhw4bB2NgYenp68PHxQVpaGh49eoRevXrBwMAAVatWxaRJk/D+/Xtx/6zhRgsXLsTcuXNhZWUFDQ0NuLq65hrT6dOn0bZtW+jq6kJLSwvNmzfHn3/+mWtMhw4dwpAhQ2BiYgItLS1MmzYNkydPBgDUqFFDPL9ZQ062bt2KDh06oGrVqtDU1IS9vT2mTp2KtLQ0hfZ9fX2ho6ODO3fuwMvLCzo6OrC0tMTEiRORnp6uUDc9PR0hISGwt7eHhoYGjI2N0bp1a0RFRYl1BEHAqlWr0KBBA2hqasLQ0BBff/017t69W6xrUhzKysr4/vvvUaVKFSxatEhhW0pKCiZNmoQaNWpATU0NFhYW8Pf3z3Fe5HI5Vq5cKfYj66Z59+7dYp3chgatXr0a9evXh46ODnR1dVG3bl3873//E7fnNTRo9+7dcHNzg5aWFnR1ddG+ffsc3+IHBQVBJpPh+vXr6Nu3L/T19WFmZoYhQ4YgOTn5E87YBzKZDGPHjsWPP/4Ie3t7qKurIyIiAgBw+/Zt9OvXD6amplBXV4e9vT1++OGHHG0U9vxSxcVEgIioAsvMzERGRobCkmXevHno27cvHBwcsG3bNmzcuBGvXr1Cy5YtcePGDbFeYmIi6tSpg2XLluHgwYNYsGABkpKS0LhxYzx9+hQA4O3tjXnz5gEAfvjhB5w9exZnz56Ft7d3seKeNm0a7t+/jx9//BF79uyBqakpfvnlF3To0AF6enqIiIjAtm3bYGRkBE9Pz2InAwDwzTffQF9fH1u2bMGMGTPw66+/YtiwYfD29kb9+vXx+++/Y9CgQViyZAlWrlyZY//vv/8eBw4cwLJly/DLL79ASUkJnTp1UrixO3HiBNq0aYPk5GT8/PPP2Lx5M3R1ddGlSxds3bo1R5tDhgyBqqoqNm7ciN9//x2jRo3Ct99+CwDYsWOHeH5dXFwAfLhx8/Lyws8//4wDBw7A398f27ZtQ5cuXXK0/f79e3Tt2hVt27bFH3/8gSFDhmDp0qVYsGCBWCcjIwOdOnXC7Nmz0blzZ+zcuRPh4eFo3rw57t+/L9YbMWIE/P390a5dO+zatQurVq3C9evX0bx5c/z333/FviZFpampiXbt2iEhIQH//PMPAOD169dwd3dHREQE/Pz8sH//fkyZMgXh4eHo2rUrBEEQ9/f19cW4cePQuHFjbN26FVu2bEHXrl3zfd9ly5YtGD16NNzd3bFz507s2rUL48ePL/Am+Ndff0W3bt2gp6eHzZs34+eff8aLFy/g4eGB06dP56jfo0cP2NnZYfv27Zg6dSp+/fVXjB8/vngnKptdu3Zh9erVmDlzJg4ePCj+3W/cuDGuXbuGJUuWYO/evfD29oafnx+Cg4PFfYtyfqkCE4iIqMIJCwsTAOS6vH//Xrh//76goqIifPvttwr7vXr1SjA3Nxd69eqVZ9sZGRlCamqqoK2tLSxfvlws/+233wQAwvHjx3PsY21tLQwaNChHubu7u+Du7i6uHz9+XAAgtGrVSqFeWlqaYGRkJHTp0kWhPDMzU6hfv77QpEmTfM6GICQkJAgAhEWLFollWeco+zno3r27AED47rvvFMobNGgguLi45GizWrVqwps3b8TylJQUwcjISGjXrp1Y1qxZM8HU1FR49eqVWJaRkSE4OjoK1atXF+RyuUJMPj4+OfqwaNEiAYCQkJCQb1/lcrnw/v174cSJEwIA4fLly+K2QYMGCQCEbdu2Kezj5eUl1KlTR1zfsGGDAEBYu3Ztnsc5e/asAEBYsmSJQvmDBw8ETU1NISAgIN84iwqAMGbMmDy3T5kyRQAgnD9/XhAEQZg/f76gpKQkXLhwQaHe77//LgAQ9u3bJwiCIJw8eVIAIEyfPj3f42f/rI4dO1YwMDDId5+sz3PW34nMzEyhWrVqgpOTk5CZmSnWe/XqlWBqaio0b95cLJs1a5YAQFi4cKFCm6NHjxY0NDTEz0xhWFtbC97e3gplAAR9fX3h+fPnCuWenp5C9erVheTkZIXysWPHChoaGmL9wp5fqtj4RICIqALbsGEDLly4oLCoqKjg4MGDyMjIgI+Pj8LTAg0NDbi7uysMZUhNTcWUKVNga2sLFRUVqKioQEdHB2lpaYiLiyuVuHv06KGwHhUVhefPn2PQoEEK8crlcnTs2BEXLlwo9nCEzp07K6zb29sDQI6nGfb29grDobJ89dVXCmP4s77pP3nyJDIzM5GWlobz58/j66+/ho6OjlhPWVkZAwcOxD///IObN2/m2/+C3L17F/369YO5uTmUlZWhqqoKd3d3AMhxjWQyWY4nBc7Ozgp9279/PzQ0NDBkyJA8j7l3717IZDIMGDBA4ZqYm5ujfv36Bc6Uk/1JlfCJ3yBn33/v3r1wdHREgwYNFI7j6empMFxn//79AIAxY8YU6XhNmjTBy5cv0bdvX/zxxx/i07H83Lx5E//++y8GDhwIJaX/u8XS0dFBjx49cO7cObx+/Vphn65duyqsOzs74+3bt3j8+HGR4s1NmzZtYGhoKK6/ffsWR48exZdffgktLS2F8+bl5YW3b9+KMzQV9vxSxSbNN7OIiCoJe3v7XF8Wzhq20bhx41z3+/gmpV+/fjh69CgCAwPRuHFj6OnpQSaTwcvLC2/evCmVuKtWrZprvF9//XWe+zx//hza2tpFPpaRkZHCupqaWp7lb9++zbG/ubl5rmXv3r1DamoqXr16BUEQcvQJ+L8ZnJ49e6ZQnlvdvKSmpqJly5bQ0NDAnDlzYGdnBy0tLTx48ABfffVVjmukpaWV4+VjdXV1hb49efIE1apVU/gcZPfff/9BEASYmZnlur1mzZp57puYmIgaNWoolB0/frzI03N+LCuRyTqn//33H+7cuQNVVdVc62fduD958gTKysq5Xsf8DBw4EBkZGVi7di169OgBuVyOxo0bY86cOWjfvn2u+2Rd57w+C3K5HC9evICWlpZYbmxsrFBPXV0dAErk7172OJ49e4aMjAysXLky12FwwP+dt8KeX6rYmAgQEVVCVapUAQD8/vvvsLa2zrNecnIy9u7di1mzZmHq1KlieXp6Op4/f17o42loaOR4GRX4cLOQFcvHss9xnlVn5cqVec58ktcNaWl79OhRrmVqamrQ0dGBiooKlJSUkJSUlKPev//+CwA5zkFR5ng/duwY/v33X0RGRopPAQB80u8NmJiY4PTp05DL5XkmA1WqVIFMJsOpU6fEm9OP5VaWpVq1arhw4YJCWZ06dYod75s3b3DkyBHUqlUL1atXF+PT1NTE+vXr84wf+NDXzMxMPHr0qEgJGAAMHjwYgwcPRlpaGk6ePIlZs2ahc+fOuHXrVq5/r7Ju6vP6LCgpKSl8Q1/asn/ODA0NxSdVeT0hyUrgCnt+qWJjIkBEVAl5enpCRUUF8fHx+Q5DkclkEAQhx03dunXrkJmZqVCW3zeVNjY2uHLlikLZrVu3cPPmzULdMLRo0QIGBga4ceMGxo4dW2D9z2nHjh1YtGiR+C37q1evsGfPHrRs2RLKysrQ1tZG06ZNsWPHDixevFicnlEul+OXX35B9erVYWdnV+Bx8jq/WTdz2a/RmjVrit2nTp06YfPmzQgPD89zeFDnzp0RGhqKhw8folevXkVqX01NrcSmtc3MzMTYsWPx7NkzzJ8/XyG+efPmwdjYOMfTh4916tQJ8+fPx+rVqxESElKsGLS1tdGpUye8e/cO3bt3x/Xr13NNBOrUqQMLCwv8+uuvmDRpknjt0tLSsH37dnEmobKipaWF1q1b49KlS3B2dhafjuWmsOeXKjYmAkRElZCNjQ1CQkIwffp03L17Fx07doShoSH+++8//PXXX9DW1kZwcDD09PTQqlUrLFq0CFWqVIGNjQ1OnDiBn3/+OcePWjk6OgIAfvrpJ+jq6kJDQwM1atSAsbExBg4ciAEDBmD06NHo0aMH7t27h4ULF8LExKRQ8ero6GDlypUYNGgQnj9/jq+//hqmpqZ48uQJLl++jCdPnmD16tUlfZoKRVlZGe3bt8eECRMgl8uxYMECpKSkKMywMn/+fLRv3x6tW7fGpEmToKamhlWrVuHatWvYvHlzoZ4AODk5AQCWL1+OQYMGQVVVFXXq1EHz5s1haGiIkSNHYtasWVBVVcWmTZtw+fLlYvepb9++CAsLw8iRI3Hz5k20bt0acrkc58+fh729Pfr06YMWLVpg+PDhGDx4MKKjo9GqVStoa2sjKSkJp0+fhpOTU4nP6f/ff//h3LlzEAQBr169En9Q7PLlyxg/fjyGDRsm1vX398f27dvRqlUrjB8/Hs7OzpDL5bh//z4OHTqEiRMnomnTpmjZsiUGDhyIOXPm4L///kPnzp2hrq6OS5cuQUtLS5ytKbthw4ZBU1MTLVq0QNWqVfHo0SPMnz8f+vr6+Q65W7hwIfr374/OnTtjxIgRSE9Px6JFi/Dy5UuEhoaW6PkqjuXLl+OLL75Ay5YtMWrUKNjY2ODVq1e4c+cO9uzZg2PHjgEo/PmlCq4s31QmIqLiyZp9JvuMHtnt2rVLaN26taCnpyeoq6sL1tbWwtdffy0cOXJErPPPP/8IPXr0EAwNDQVdXV2hY8eOwrVr13KdCWjZsmVCjRo1BGVlZQGAEBYWJgjCh5lsFi5cKNSsWVPQ0NAQXF1dhWPHjuU5a9Bvv/2Wa7wnTpwQvL29BSMjI0FVVVWwsLAQvL2986yfJb9Zg7Kfo6zZWp48eaJQPmjQIEFbWztHmwsWLBCCg4OF6tWrC2pqakLDhg2FgwcP5ojh1KlTQps2bQRtbW1BU1NTaNasmbBnzx6FOgVdt2nTpgnVqlUTlJSUFGajiYqKEtzc3AQtLS3BxMRE+Oabb4SLFy8qXIPc+pC9zx978+aNMHPmTKF27dqCmpqaYGxsLLRp00aIiopSqLd+/XqhadOmYr9q1aol+Pj4CNHR0bn2objw0cxXSkpKgp6enuDk5CQMHz5cOHv2bK77pKamCjNmzBDq1KkjqKmpCfr6+oKTk5Mwfvx44dGjR2K9zMxMYenSpYKjo6NYz83NTeH6ZP+sRkRECK1btxbMzMwENTU1oVq1akKvXr2EK1euiHWyzxqUZdeuXULTpk0FDQ0NQVtbW2jbtq1w5swZhTp5fQ6zPiMFzR71sbxmDcprFqaEhARhyJAhgoWFhaCqqiqYmJgIzZs3F+bMmaNQr7DnlyoumSBwIlgiIqLssl54XbRoESZNmlTW4RARlThOH0pEREREJEFMBIiIiIiIJIhDg4iIiIiIJIhPBIiIiIiIJIiJABERERGRBDERICIiIiKSIP6gGBHlSi6X499//4Wurm6hfgyJiIiIyp7w/3+Qr1q1alBSyv87fyYCRJSrf//9F5aWlmUdBhERERXDgwcPUL169XzrMBEgolzp6uoC+PA/Ej09vTKOhoiIiAojJSUFlpaW4r/j+WEiQES5yhoOpKenx0SAiIioginMsF6+LExEREREJEFMBIiIiIiIJIiJABERERGRBDERICIiIiKSICYCREREREQSxESAiIiIiEiCmAgQEREREUkQEwEiIiIiIgliIkBEREREJEFMBIiIiIiIJIiJABERERGRBDERICIiIiKSICYCREREREQSxESAiIiIiEiCmAgQUb4cZx2EzdQ/yzoMIiIiKmFMBIiIiIiIJIiJABERERGRBDERICIiIiKSICYCREREREQSxESAiIiIiEiCmAgQEREREUkQEwEiIiIiIgliIkBEREREJEFMBIiIiIiIJIiJABERERGRBDERICIiIiKSICYCREREREQSxESAiIiIiEiCmAgQEREREUkQEwGiciwxMREymQyxsbEAgMjISMhkMrx8+bJM4yIiIqKKj4kASZqHhwf8/f1L/TjFvYG3tLREUlISHB0dSycwIiIikiwmAvTZvXv3rqxDKDGCICAjI6PU2ldWVoa5uTlUVFRKrM3KdP6JiIio+JgIUKnz8PDA2LFjMWHCBFSpUgXt27fHjRs34OXlBR0dHZiZmWHgwIF4+vRpjn3Gjh0LAwMDGBsbY8aMGRAEQazz7t07BAQEwMLCAtra2mjatCkiIyPF7c+ePUPfvn1RvXp1aGlpwcnJCZs3bxa3+/r64sSJE1i+fDlkMhlkMhkSExPz7UvWN/sHDx6Eq6sr1NXVcerUKQiCgIULF6JmzZrQ1NRE/fr18fvvvwP4MLyndevWAABDQ0PIZDL4+voCAA4cOIAvvvhC7GPnzp0RHx8vHi/70KDcREVFoVWrVtDU1ISlpSX8/PyQlpYmbrexscGcOXPg6+sLfX19DBs2LN8+EhERkTQwEaDPIiIiAioqKjhz5gxCQ0Ph7u6OBg0aIDo6GgcOHMB///2HXr165brP+fPnsWLFCixduhTr1q0Ttw8ePBhnzpzBli1bcOXKFfTs2RMdO3bE7du3AQBv375Fo0aNsHfvXly7dg3Dhw/HwIEDcf78eQDA8uXL4ebmhmHDhiEpKQlJSUmwtLQsVH8CAgIwf/58xMXFwdnZGTNmzEBYWBhWr16N69evY/z48RgwYABOnDgBS0tLbN++HQBw8+ZNJCUlYfny5QCAtLQ0TJgwARcuXMDRo0ehpKSEL7/8EnK5vFBxXL16FZ6envjqq69w5coVbN26FadPn8bYsWMV6i1atAiOjo6IiYlBYGBgrm2lp6cjJSVFYSEiIqJKTCAqZe7u7kKDBg3E9cDAQKFDhw4KdR48eCAAEG7evCnuY29vL8jlcrHOlClTBHt7e0EQBOHOnTuCTCYTHj58qNBO27ZthWnTpuUZi5eXlzBx4kSF2MaNG1fovhw/flwAIOzatUssS01NFTQ0NISoqCiFukOHDhX69u2rsN+LFy/ybf/x48cCAOHq1auCIAhCQkKCAEC4dOlSru0MHDhQGD58uEIbp06dEpSUlIQ3b94IgiAI1tbWQvfu3Qvs26xZswQAORZL/22C9ZS9Be5PREREZS85OVkAICQnJxdYt+QGHhPlw9XVVfxzTEwMjh8/Dh0dnRz14uPjYWdnBwBo1qwZZDKZuM3NzQ1LlixBZmYmLl68CEEQxLpZ0tPTYWxsDADIzMxEaGgotm7diocPHyI9PR3p6enQ1tYu0f7cuHEDb9++Rfv27RXqvHv3Dg0bNsy3nfj4eAQGBuLcuXN4+vSp+CTg/v37hXpBOCYmBnfu3MGmTZvEMkEQIJfLkZCQAHt7+xzx5mXatGmYMGGCuJ6SklLoJyRERERU8TARoM/i45tvuVyOLl26YMGCBTnqVa1atVDtyeVyKCsrIyYmBsrKygrbshKMJUuWYOnSpVi2bBmcnJygra0Nf3//EnlZNnt/AODPP/+EhYWFQj11dfV82+nSpQssLS2xdu1aVKtWDXK5HI6OjoWOUS6XY8SIEfDz88uxzcrKKtd486Kurl5gvERERFR5MBGgz87FxQXbt2+HjY1NvrPhnDt3Lsd67dq1oaysjIYNGyIzMxOPHz9Gy5Ytc93/1KlT6NatGwYMGADgw03z7du3xW/JAUBNTQ2ZmZmf1B8HBweoq6vj/v37cHd3z7WOmpoaACgc69mzZ4iLi8OaNWvEPpw+fbpIx3ZxccH169dha2tbzOiJiIhIqviyMH12Y8aMwfPnz9G3b1/89ddfuHv3Lg4dOoQhQ4Yo3Cg/ePAAEyZMwM2bN7F582asXLkS48aNAwDY2dmhf//+8PHxwY4dO5CQkIALFy5gwYIF2LdvHwDA1tYWhw8fRlRUFOLi4jBixAg8evRIIRYbGxucP38eiYmJCkNzikJXVxeTJk3C+PHjERERgfj4eFy6dAk//PADIiIiAADW1taQyWTYu3cvnjx5gtTUVBgaGsLY2Bg//fQT7ty5g2PHjikMzSmMKVOm4OzZsxgzZgxiY2Nx+/Zt7N69G99++22R+0FERETSwkSAPrtq1arhzJkzyMzMhKenJxwdHTFu3Djo6+tDSen/PpI+Pj548+YNmjRpgjFjxuDbb7/F8OHDxe1hYWHw8fHBxIkTUadOHXTt2hXnz58Xx7UHBgbCxcUFnp6e8PDwgLm5Obp3764Qy6RJk6CsrAwHBweYmJjg/v37xerT7NmzMXPmTMyfPx/29vbw9PTEnj17UKNGDQCAhYUFgoODMXXqVJiZmWHs2LFQUlLCli1bEBMTA0dHR4wfPx6LFi0q0nGdnZ1x4sQJ3L59Gy1btkTDhg0RGBhY6CFWREREJF0yQfhoYnaicsLDwwMNGjTAsmXLyjoUyUpJSYG+vj4s/bdBSV0LiaHeZR0SERERFSDr3+/k5GTo6enlW5dPBIiIiIiIJIiJANFHRo4cCR0dnVyXkSNHlnV4RERERCWGQ4OIPvL48eM8f1FXT08PpqamnzmissOhQURERBVPUYYGcfpQoo+YmppK6mafiIiIpItDg4iIiIiIJIiJABERERGRBDERICIiIiKSICYCREREREQSxESAiIiIiEiCmAgQEREREUkQEwEiIiIiIgliIkBEREREJEH8QTEiyte1YM8Cf5mQiIiIKh4+ESAiIiIikiAmAkREREREEsREgIiIiIhIgpgIEBERERFJEBMBIiIiIiIJYiJARERERCRBTASIiIiIiCSIiQARERERkQQxESAiIiIikiD+sjAR5ctx1kEoqWuVdRg5JIZ6l3UIREREFRqfCBARERERSRATASIiIiIiCWIiQEREREQkQUwEiIiIiIgkiIkAEREREZEEMREgIiIiIpIgJgJERERERBLERICIiIiISIKYCBARERERSRATASIiIiIiCWIiQEREREQkQUwEiIiIiIgkiIkAEREREZEEMREgIiIiIpKgCpkIeHh4wN/fHwBgY2ODZcuWlWk8pe3169fo0aMH9PT0IJPJ8PLly7IOSRQeHg4DA4OyDqPS+/gzT0RERFQSKmQi8LELFy5g+PDhhapbUZOGiIgInDp1ClFRUUhKSoK+vn5Zh1RqinrDGxkZWe6So0+RV3927NiB2bNnl01QREREVCmplHUAn8rExKSsQyh18fHxsLe3h6OjY5513r17BzU1tc8YVcVS0c+PkZFRWYdARERElUy5fyKQlpYGHx8f6OjooGrVqliyZInC9uzf8gcFBcHKygrq6uqoVq0a/Pz8AHz4pvnevXsYP348ZDIZZDIZAODZs2fo27cvqlevDi0tLTg5OWHz5s0Kx/Dw8ICfnx8CAgJgZGQEc3NzBAUFKdR5+fIlhg8fDjMzM2hoaMDR0RF79+4Vt0dFRaFVq1bQ1NSEpaUl/Pz8kJaWVmD/PTw8sGTJEpw8eRIymQweHh5iv+fMmQNfX1/o6+tj2LBhAIDt27ejXr16UFdXh42NTa7na86cOeI5tba2xh9//IEnT56gW7du0NHRgZOTE6KjowuMLTfx8fHo1q0bzMzMoKOjg8aNG+PIkSMKdVatWoXatWtDQ0MDZmZm+PrrrwEAvr6+OHHiBJYvXy5eo8TExDyPlZiYiNatWwMADA0NIZPJ4OvrK563sWPHYsKECahSpQrat28PAPjuu+/g5OQEbW1tWFpaYvTo0UhNTRXbzBrqdPDgQdjb20NHRwcdO3ZEUlKSWCcyMhJNmjSBtrY2DAwM0KJFC9y7d6/Q/U9PT0dAQAAsLS2hrq6O2rVr4+effy6wPx8/KXnx4gV8fHxgaGgILS0tdOrUCbdv3y5SP4iIiEjayn0iMHnyZBw/fhw7d+7EoUOHEBkZiZiYmFzr/v7771i6dCnWrFmD27dvY9euXXBycgLwYWhF9erVERISgqSkJPGG6O3bt2jUqBH27t2La9euYfjw4Rg4cCDOnz+v0HZERAS0tbVx/vx5LFy4ECEhITh8+DAAQC6Xo1OnToiKisIvv/yCGzduIDQ0FMrKygCAq1evwtPTE1999RWuXLmCrVu34vTp0xg7dmyB/d+xYweGDRsGNzc3JCUlYceOHeK2RYsWwdHRETExMQgMDERMTAx69eqFPn364OrVqwgKCkJgYCDCw8MV2ly6dClatGiBS5cuwdvbGwMHDoSPjw8GDBiAixcvwtbWFj4+PhAEoXAX6SOpqanw8vLCkSNHcOnSJXh6eqJLly64f/8+ACA6Ohp+fn4ICQnBzZs3ceDAAbRq1QoAsHz5cri5uWHYsGHiNbK0tMzzWJaWlti+fTsA4ObNm0hKSsLy5cvF7REREVBRUcGZM2ewZs0aAICSkhJWrFiBa9euISIiAseOHUNAQIBCu69fv8bixYuxceNGnDx5Evfv38ekSZMAABkZGejevTvc3d1x5coVnD17FsOHDxcTy4L6DwA+Pj7YsmULVqxYgbi4OPz444/Q0dEpsD8f8/X1RXR0NHbv3o2zZ89CEAR4eXnh/fv3hepHbtLT05GSkqKwEBERUeVVrocGpaam4ueff8aGDRvEb3QjIiJQvXr1XOvfv38f5ubmaNeuHVRVVWFlZYUmTZoA+DC0QllZGbq6ujA3Nxf3sbCwULg5+vbbb3HgwAH89ttvaNq0qVju7OyMWbNmAQBq166N77//HkePHkX79u1x5MgR/PXXX4iLi4OdnR0AoGbNmuK+ixYtQr9+/cRvdGvXro0VK1bA3d0dq1evhoaGRp7nwMjICFpaWlBTU1OIGwDatGmjEHv//v3Rtm1bBAYGAgDs7Oxw48YNLFq0SPxmGQC8vLwwYsQIAMDMmTOxevVqNG7cGD179gQATJkyBW5ubvjvv/9yHLMg9evXR/369cX1OXPmYOfOndi9ezfGjh2L+/fvQ1tbG507d4auri6sra3RsGFDAIC+vj7U1NSgpaVVqOMqKyuLQ2ZMTU1zvLRsa2uLhQsXKpR9/K16jRo1MHv2bIwaNQqrVq0Sy9+/f48ff/wRtWrVAgCMHTsWISEhAICUlBQkJyejc+fO4nZ7e/tC9//WrVvYtm0bDh8+jHbt2gFQ/Kzk158st2/fxu7du3HmzBk0b94cALBp0yZYWlpi165d4nXMrx+5mT9/PoKDg/PcTkRERJVLuX4iEB8fj3fv3sHNzU0sMzIyQp06dXKt37NnT7x58wY1a9bEsGHDsHPnTmRkZOR7jMzMTMydOxfOzs4wNjaGjo4ODh06pPANLvAhEfhY1apV8fjxYwBAbGwsqlevLiYB2cXExCA8PBw6Ojri4unpCblcjoSEhALPQ15cXV0V1uPi4tCiRQuFshYtWuD27dvIzMzMtS9mZmYAID45+bgsq39FkZaWhoCAADg4OMDAwAA6Ojr4+++/xfPZvn17WFtbo2bNmhg4cCA2bdqE169fF/k4hZH9/ADA8ePH0b59e1hYWEBXVxc+Pj549uyZwjAtLS0t8eYZULzWRkZG8PX1Fb/pX758ucJwm4L6HxsbC2VlZbi7uxe7X3FxcVBRUVFIVI2NjVGnTh3ExcUVqh+5mTZtGpKTk8XlwYMHxY6RiIiIyr9ynQgUdWiKpaUlbt68iR9++AGampoYPXo0WrVqpTBcIrslS5Zg6dKlCAgIwLFjxxAbGwtPT0+8e/dOoZ6qqqrCukwmg1wuBwBoamrmG5dcLseIESMQGxsrLpcvX8bt27cVbtSKSltbW2FdEARxiMrHZdl93Jes+rmVZfWvKCZPnozt27dj7ty5OHXqFGJjY+Hk5CSeT11dXVy8eBGbN29G1apVMXPmTNSvX79UZv3Jfn7u3bsHLy8vODo6Yvv27YiJicEPP/wAAAqfkdyu9cfnMSwsDGfPnkXz5s2xdetW2NnZ4dy5c4Xqf0GflcLI6+9F9utfUD+yU1dXh56ensJCRERElVe5TgRsbW2hqqoq3mQBH16SvHXrVp77aGpqomvXrlixYgUiIyNx9uxZXL16FQCgpqam8M04AJw6dQrdunXDgAEDUL9+fdSsWVPhpcvCcHZ2xj///JNnXC4uLrh+/TpsbW1zLCU5k42DgwNOnz6tUBYVFQU7OzvxfYXSdurUKfj6+uLLL7+Ek5MTzM3Nc7zwq6Kignbt2mHhwoW4cuUKEhMTcezYMQC5X6P8ZJ2/wuwTHR2NjIwMLFmyBM2aNYOdnR3+/fffwnfuIw0bNsS0adMQFRUFR0dH/PrrrwAK7r+TkxPkcjlOnDhR7P44ODggIyND4T2WZ8+e4datWwrDlIiIiIjyU64TAR0dHQwdOhSTJ0/G0aNHce3aNfj6+kJJKfeww8PD8fPPP+PatWu4e/cuNm7cCE1NTVhbWwP4MGPOyZMn8fDhQzx9+hTAh2Tj8OHDiIqKQlxcHEaMGIFHjx4VKU53d3e0atUKPXr0wOHDh5GQkID9+/fjwIEDAD6MuT979izGjBmD2NhYcYz3t99++wlnJ6eJEyfi6NGjmD17Nm7duoWIiAh8//33+b4gWtJsbW2xY8cO8alHv379FJ4s7N27FytWrEBsbCzu3buHDRs2QC6Xi8O9bGxscP78eSQmJuLp06cFPpWwtraGTCbD3r178eTJE4UZgLKrVasWMjIysHLlSvHz8eOPPxapfwkJCZg2bRrOnj2Le/fu4dChQwo34AX138bGBoMGDcKQIUOwa9cuJCQkIDIyEtu2bSt0f2rXro1u3bph2LBhOH36NC5fvowBAwbAwsIC3bp1K1J/iIiISLrKdSIAfHjRtlWrVujatSvatWuHL774Ao0aNcq1roGBAdauXYsWLVrA2dkZR48exZ49e2BsbAwACAkJQWJiImrVqiX+/kBgYCBcXFzg6ekJDw8PmJubo3v37kWOc/v27WjcuDH69u0LBwcHBAQEiN/qOjs748SJE7h9+zZatmyJhg0bIjAwEFWrVi3eScmDi4sLtm3bhi1btsDR0REzZ85ESEiIwovCpW3p0qUwNDRE8+bN0aVLF3h6esLFxUXcbmBggB07dqBNmzawt7fHjz/+iM2bN6NevXoAgEmTJkFZWRkODg4wMTHJ8a5GdhYWFggODsbUqVNhZmaW70xMDRo0wHfffYcFCxbA0dERmzZtwvz584vUPy0tLfz999/o0aMH7OzsMHz4cIwdO1Z8+bqg/gPA6tWr8fXXX2P06NGoW7cuhg0bJr6jUNj+hIWFoVGjRujcuTPc3NwgCAL27duXYzgQERERUV5kQnHmiCSiSi8lJQX6+vqw9N8GJXWtsg4nh8RQ77IOgYiIqNzJ+vc7OTm5wPf9yv0TASIiIiIiKnlMBMrYqVOnFKYVzb6UtU6dOuUZ27x580r9+CNHjszz+CNHjiz14xMRERFVVhwaVMbevHmDhw8f5rnd1tb2M0aT08OHD/HmzZtctxkZGYk/gFVaHj9+nOcv3Orp6cHU1LRUjy9lHBpERERU8RRlaFC5/mVhKdDU1Czzm/38WFhYlOnxTU1NebNPREREVAo4NIiIiIiISIKYCBARERERSRATASIiIiIiCWIiQEREREQkQUwEiIiIiIgkiIkAEREREZEEMREgIiIiIpIgJgJERERERBLERICIiIiISIL4y8JElK9rwZ4F/kQ5ERERVTx8IkBEREREJEFMBIiIiIiIJIiJABERERGRBDERICIiIiKSICYCREREREQSxESAiIiIiEiCmAgQEREREUkQEwEiIiIiIgliIkBEREREJEH8ZWEiypfjrINQUtcq6zAqhMRQ77IOgYiIqND4RICIiIiISIKYCBARERERSRATASIiIiIiCWIiQEREREQkQUwEiIiIiIgkiIkAEREREZEEMREgIiIiIpIgJgJERERERBLERICIiIiISIKYCBARERERSRATASIiIiIiCWIiQEREREQkQUwEiIiIiIgkiIkAUSkIDw+HgYFBibRlY2ODZcuW5bk9MTERMpkMsbGxJXI8IiIikgaVsg6AiD6NpaUlkpKSUKVKlbIOhYiIiCoQJgJEFZyysjLMzc3LOgwiIiKqYDg0iCQht+E1DRo0QFBQEAAgKCgIVlZWUFdXR7Vq1eDn5yfWe/fuHQICAmBhYQFtbW00bdoUkZGRCm2Fh4fDysoKWlpa+PLLL/Hs2bMixbd79264urpCQ0MDVapUwVdffaWw/fXr1xgyZAh0dXVhZWWFn376SdyWfWhQZGQkZDIZ/vzzT9SvXx8aGhpo2rQprl69WqSYiIiIqHJjIkCS9/vvv2Pp0qVYs2YNbt++jV27dsHJyUncPnjwYJw5cwZbtmzBlStX0LNnT3Ts2BG3b98GAJw/fx5DhgzB6NGjERsbi9atW2POnDmFPv6ff/6Jr776Ct7e3rh06RKOHj0KV1dXhTpLliyBq6srLl26hNGjR2PUqFH4+++/82138uTJWLx4MS5cuABTU1N07doV79+/L8KZISIiosqMQ4NI8u7fvw9zc3O0a9cOqqqqsLKyQpMmTQAA8fHx2Lx5M/755x9Uq1YNADBp0iQcOHAAYWFhmDdvHpYvXw5PT09MnToVAGBnZ4eoqCgcOHCgUMefO3cu+vTpg+DgYLGsfv36CnW8vLwwevRoAMCUKVOwdOlSREZGom7dunm2O2vWLLRv3x4AEBERgerVq2Pnzp3o1atXrvXT09ORnp4urqekpBQqfiIiIqqY+ESAJK9nz5548+YNatasiWHDhmHnzp3IyMgAAFy8eBGCIMDOzg46OjricuLECcTHxwMA4uLi4ObmptBm9vX8xMbGom3btvnWcXZ2Fv8sk8lgbm6Ox48f57vPxzEYGRmhTp06iIuLy7P+/Pnzoa+vLy6WlpaF7AERERFVRHwiQJKgpKQEQRAUyrKGyVhaWuLmzZs4fPgwjhw5gtGjR2PRokU4ceIE5HI5lJWVERMTA2VlZYX9dXR0ACBHu0WlqalZYB1VVVWFdZlMBrlcXuRjyWSyPLdNmzYNEyZMENdTUlKYDBAREVViTARIEkxMTJCUlCSup6SkICEhQVzX1NRE165d0bVrV4wZMwZ169bF1atX0bBhQ2RmZuLx48do2bJlrm07ODjg3LlzCmXZ1/Pj7OyMo0ePYvDgwUXsVf7OnTsHKysrAMCLFy9w69atfIcSqaurQ11dvURjICIiovKLiQBJQps2bRAeHo4uXbrA0NAQgYGB4jf84eHhyMzMRNOmTaGlpYWNGzdCU1MT1tbWMDY2Rv/+/eHj44MlS5agYcOGePr0KY4dOwYnJyd4eXnBz88PzZs3x8KFC9G9e3ccOnSo0O8HAB/G8rdt2xa1atVCnz59kJGRgf379yMgIOCT+hwSEgJjY2OYmZlh+vTpqFKlCrp37/5JbRIREVHlwXcESBKmTZuGVq1aoXPnzvDy8kL37t1Rq1YtAICBgQHWrl2LFi1aiN/O79mzB8bGxgCAsLAw+Pj4YOLEiahTpw66du2K8+fPi8NmmjVrhnXr1mHlypVo0KABDh06hBkzZhQ6Ng8PD/z222/YvXs3GjRogDZt2uD8+fOf3OfQ0FCMGzcOjRo1QlJSEnbv3g01NbVPbpeIiIgqB5nwqQOciahciYyMROvWrfHixQsYGBgUu52UlJQPLw37b4OSulbJBViJJYZ6l3UIREQkcVn/ficnJ0NPTy/funwiQEREREQkQUwEiEpZvXr1FKYe/XjZtGlTWYdHREREEsWXhYlK2b59+/L8RV8zM7MSP56Hh8cnT2lKRERElR8TAaJSZm1tXdYhEBEREeXAoUFERERERBLERICIiIiISIKYCBARERERSRATASIiIiIiCWIiQEREREQkQUwEiIiIiIgkiIkAEREREZEEMREgIiIiIpIgJgJERERERBLEXxYmonxdC/aEnp5eWYdBREREJYxPBIiIiIiIJIiJABERERGRBDERICIiIiKSICYCREREREQSxESAiIiIiEiCmAgQEREREUlQsROBjRs3okWLFqhWrRru3bsHAFi2bBn++OOPEguOiIiIiIhKR7ESgdWrV2PChAnw8vLCy5cvkZmZCQAwMDDAsmXLSjI+IiIiIiIqBcVKBFauXIm1a9di+vTpUFZWFstdXV1x9erVEguOiIiIiIhKR7F+WTghIQENGzbMUa6uro60tLRPDoqIyg/HWQehpK5V1mFQCUkM9S7rEIiIqJwo1hOBGjVqIDY2Nkf5/v374eDg8KkxERERERFRKSvWE4HJkydjzJgxePv2LQRBwF9//YXNmzdj/vz5WLduXUnHSEREREREJaxYicDgwYORkZGBgIAAvH79Gv369YOFhQWWL1+OPn36lHSMRERERERUwoqcCGRkZGDTpk3o0qULhg0bhqdPn0Iul8PU1LQ04iMiIiIiolJQ5HcEVFRUMGrUKKSnpwMAqlSpwiSAiIiIiKiCKdbLwk2bNsWlS5dKOhYiIiIiIvpMivWOwOjRozFx4kT8888/aNSoEbS1tRW2Ozs7l0hwRERERERUOoqVCPTu3RsA4OfnJ5bJZDIIggCZTCb+0jAREREREZVPxf5BMSIiIiIiqriKlQhYW1uXdBxERERERPQZFSsR2LBhQ77bfXx8ihUMERERERF9HsVKBMaNG6ew/v79e7x+/RpqamrQ0tJiIkBEREREVM4Va/rQFy9eKCypqam4efMmvvjiC2zevLmkYyQiIiIiohJWrEQgN7Vr10ZoaGiOpwVERERERFT+lFgiAADKysr4999/S7JJkoDw8HAYGBiUWHu+vr7o3r17ibVXlmQyGXbt2lXWYRAREVElVKx3BHbv3q2wLggCkpKS8P3336NFixYlEhhRcS1fvhyCIJR1GERERETlWrESgezftspkMpiYmKBNmzZYsmRJScRFVGz6+vplHUK59u7dO6ipqZV1GERERFTGijU0SC6XKyyZmZl49OgRfv31V1StWrWkY6TPzMbGBsuWLVMoa9CgAYKCggAAQUFBsLKygrq6OqpVq6bwC9Pv3r1DQEAALCwsoK2tjaZNmyIyMlKhrfDwcFhZWUFLSwtffvklnj17VujYgoKC0KBBA6xZswaWlpbQ0tJCz5498fLlS7FO9qFBHh4e8PPzQ0BAAIyMjGBubi72pTBkMhnWrVuHL7/8ElpaWqhdu7bCU7Hchjbt2rULMpksR9zr16+HlZUVdHR0MGrUKGRmZmLhwoUwNzeHqakp5s6dm+P4SUlJ6NSpEzQ1NVGjRg389ttvCtsfPnyI3r17w9DQEMbGxujWrRsSExNznI/58+ejWrVqsLOzK3TfiYiIqPIqViIQEhKC169f5yh/8+YNQkJCPjkoKr9+//13LF26FGvWrMHt27exa9cuODk5idsHDx6MM2fOYMuWLbhy5Qp69uyJjh074vbt2wCA8+fPY8iQIRg9ejRiY2PRunVrzJkzp0gx3LlzB9u2bcOePXtw4MABxMbGYsyYMfnuExERAW1tbZw/fx4LFy5ESEgIDh8+XOhjBgcHo1evXrhy5Qq8vLzQv39/PH/+vEhxx8fHY//+/Thw4AA2b96M9evXw9vbG//88w9OnDiBBQsWYMaMGTh37pzCfoGBgejRowcuX76MAQMGoG/fvoiLiwMAvH79Gq1bt4aOjg5OnjyJ06dPQ0dHBx07dsS7d+/ENo4ePYq4uDgcPnwYe/fuzTW+9PR0pKSkKCxERERUeRUrEQgODkZqamqO8tevXyM4OPiTg6Ly6/79+zA3N0e7du1gZWWFJk2aYNiwYQA+3Ohu3rwZv/32G1q2bIlatWph0qRJ+OKLLxAWFgbgw/h9T09PTJ06FXZ2dvDz84Onp2eRYnj79i0iIiLQoEEDtGrVCitXrsSWLVvw6NGjPPdxdnbGrFmzULt2bfj4+MDV1RVHjx4t9DF9fX3Rt29f2NraYt68eUhLS8Nff/1VpLjlcjnWr18PBwcHdOnSBa1bt8bNmzexbNky1KlTB4MHD0adOnVyPEHp2bMnvvnmG9jZ2WH27NlwdXXFypUrAQBbtmyBkpIS1q1bBycnJ9jb2yMsLAz3799XaEdbWxvr1q1DvXr14OjomGt88+fPh76+vrhYWloWqX9ERERUsRQrERAEQWHYQ5bLly/DyMjok4Oi8qtnz5548+YNatasiWHDhmHnzp3IyMgAAFy8eBGCIMDOzg46OjricuLECcTHxwMA4uLi4ObmptBm9vWCWFlZoXr16gr7y+Vy3Lx5M899nJ2dFdarVq2Kx48fF/qYH++vra0NXV3dIu0PfBhypaurK66bmZnBwcEBSkpKCmXZ283tfGU9EYiJicGdO3egq6srnm8jIyO8fftWPOcA4OTkVOB7AdOmTUNycrK4PHjwoEj9IyIiooqlSC8LGxoaQiaTQSaTwc7OTiEZyMzMRGpqKkaOHFniQdLnpaSklGPWnffv3wMALC0tcfPmTRw+fBhHjhzB6NGjsWjRIpw4cQJyuRzKysqIiYmBsrKywv46OjoAUCqz+WR9DnNLTrOoqqrm2Eculxf6GPntn9/5KqiN4saV1Ve5XI5GjRph06ZNOeqYmJiIf9bW1i6wTXV1dairqxdYj4iIiCqHIiUCy5YtgyAIGDJkCIKDgxVmZ1FTU4ONjU2Rv92l8sfExARJSUniekpKChISEsR1TU1NdO3aFV27dsWYMWNQt25dXL16FQ0bNkRmZiYeP36Mli1b5tq2g4NDjjHw2dcLcv/+ffz777+oVq0aAODs2bNQUlIqs5dgTUxM8OrVK6SlpYk33LGxsSXW/rlz5+Dj46Ow3rBhQwCAi4sLtm7dClNTU+jp6ZXYMYmIiKjyK1IiMGjQIABAjRo10Lx58xzfZlLl0KZNG4SHh6NLly4wNDREYGCg+A1/eHg4MjMz0bRpU2hpaWHjxo3Q1NSEtbU1jI2N0b9/f/j4+GDJkiVo2LAhnj59imPHjsHJyQleXl7w8/ND8+bNsXDhQnTv3h2HDh3CgQMHihSfhoYGBg0ahMWLFyMlJQV+fn7o1asXzM3NS+N0FCjrXPzvf//Dt99+i7/++gvh4eEl1v5vv/0GV1dXfPHFF9i0aRP++usv/PzzzwCA/v37Y9GiRejWrRtCQkJQvXp13L9/Hzt27MDkyZMVhlARERERfaxY7wi4u7uLScCbN28400glM23aNLRq1QqdO3eGl5cXunfvjlq1agEADAwMsHbtWrRo0QLOzs44evQo9uzZA2NjYwBAWFgYfHx8MHHiRNSpUwddu3bF+fPnxRdPmzVrhnXr1mHlypVo0KABDh06hBkzZhQpPltbW3z11Vfw8vJChw4d4OjoiFWrVpXsSSgCIyMj/PLLL9i3bx+cnJywefPmIk1PWpDg4GBs2bIFzs7OiIiIwKZNm+Dg4AAA0NLSwsmTJ2FlZYWvvvoK9vb2GDJkCN68ecMnBERERJQvmVCMQduvX79GQEAAtm3blusc8JmZmSUSHFF2QUFB2LVrV4kOvaHcpaSkfJg9yH8blNS1yjocKiGJod5lHQIREZWirH+/k5OTC/xSsFhPBCZPnoxjx45h1apVUFdXx7p16xAcHIxq1aphw4YNxQqaiIiIiIg+n2IlAnv27MGqVavw9ddfQ0VFBS1btsSMGTMwb968XGcvISqsevXqKUw9+vFSGp+tTZs25Xm8evXqlfjxiIiIiMqLIr0snOX58+eoUaMGAEBPT0/8hdUvvvgCo0aNKrnoSHL27duX69SbwIc59nV1dUt0/H3Xrl3RtGnTXLfxZXgiIiKqzIqVCNSsWROJiYmwtraGg4MDtm3bhiZNmmDPnj0wMDAo4RBJSqytrT/r8XR1dRV+5IuIiIhIKoo1NGjw4MG4fPkygA8zzGS9KzB+/HhMnjy5RAMkIiIiIqKSV6wnAuPHjxf/3Lp1a/z999+Ijo5GrVq1UL9+/RILjoiIiIiISkexEoGPvX37FlZWVrCysiqJeIiIiIiI6DMo1tCgzMxMzJ49GxYWFtDR0cHdu3cBAIGBgeIvnhIRERERUflVrERg7ty5CA8Px8KFC6GmpiaWOzk5Yd26dSUWHBERERERlY5iJQIbNmzATz/9hP79+0NZWVksd3Z2xt9//11iwRERERERUekoViLw8OFD2Nra5iiXy+V5zgFPRERERETlR7ESgXr16uHUqVM5yn/77Tc0bNjwk4MiIiIiIqLSVaxZg2bNmoWBAwfi4cOHkMvl2LFjB27evIkNGzZg7969JR0jERERERGVMJkgCEJhK9+9exc1atSATCbDwYMHMW/ePMTExEAul8PFxQUzZ85Ehw4dSjNeIvpMUlJSoK+vj+TkZOjp6ZV1OERERFQIRfn3u0hPBGrXro2kpCSYmprC09MT69evx507d2Bubv5JARMRERER0edVpHcEsj882L9/P16/fl2iARERERERUekr1svCWYowqoiIiIiIiMqRIiUCMpkMMpksRxkREREREVUsRXpHQBAE+Pr6Ql1dHQDw9u1bjBw5Etra2gr1duzYUXIREhERERFRiStSIjBo0CCF9QEDBpRoMERERERE9HkUKREICwsrrTiIiIiIiOgz+qSXhYmIiIiIqGJiIkBEREREJEFFGhpERNLjOOsglNS1yjoMKkWJod5lHQIREZUBPhEgIiIiIpIgJgJERERERBLERICIiIiISIKYCBARERERSRATASIiIiIiCWIiQEREREQkQUwEiIiIiIgkiIkAEREREZEEMREgIiIiIpIgJgJERERERBLERICIiIiISIKYCBARERERSRATASIiIiIiCWIiUMl4eHjA398fAGBjY4Nly5aVaTyl7fXr1+jRowf09PQgk8nw8uXLsg5JFB4eDgMDg7IOg4iIiChXTAQqsQsXLmD48OGFqltRk4aIiAicOnUKUVFRSEpKgr6+flmHRERERFQhqJR1AFR6TExMyjqEUhcfHw97e3s4OjrmWefdu3dQU1P7jFERERERlX98IlCBpaWlwcfHBzo6OqhatSqWLFmisD37t/xBQUGwsrKCuro6qlWrBj8/PwAfhhPdu3cP48ePh0wmg0wmAwA8e/YMffv2RfXq1aGlpQUnJyds3rxZ4RgeHh7w8/NDQEAAjIyMYG5ujqCgIIU6L1++xPDhw2FmZgYNDQ04Ojpi79694vaoqCi0atUKmpqasLS0hJ+fH9LS0grsv4eHB5YsWYKTJ09CJpPBw8ND7PecOXPg6+sLfX19DBs2DACwfft21KtXD+rq6rCxscn1fM2ZM0c8p9bW1vjjjz/w5MkTdOvWDTo6OnByckJ0dHSBsX3s4MGDsLe3h46ODjp27IikpCRxm1wuR0hICKpXrw51dXU0aNAABw4cELcnJiZCJpNh27ZtaNmyJTQ1NdG4cWPcunULFy5cgKurq9jukydPFI4bFhYGe3t7aGhooG7duli1alWR4iYiIqLKjYlABTZ58mQcP34cO3fuxKFDhxAZGYmYmJhc6/7+++9YunQp1qxZg9u3b2PXrl1wcnICAOzYsQPVq1dHSEgIkpKSxBvVt2/folGjRti7dy+uXbuG4cOHY+DAgTh//rxC2xEREdDW1sb58+excOFChISE4PDhwwA+3Oh26tQJUVFR+OWXX3Djxg2EhoZCWVkZAHD16lV4enriq6++wpUrV7B161acPn0aY8eOLbD/O3bswLBhw+Dm5oakpCTs2LFD3LZo0SI4OjoiJiYGgYGBiImJQa9evdCnTx9cvXoVQUFBCAwMRHh4uEKbS5cuRYsWLXDp0iV4e3tj4MCB8PHxwYABA3Dx4kXY2trCx8cHgiAU6hq9fv0aixcvxsaNG3Hy5Encv38fkyZNErcvX74cS5YsweLFi3HlyhV4enqia9euuH37tkI7s2bNwowZM3Dx4kWoqKigb9++CAgIwPLly3Hq1CnEx8dj5syZYv21a9di+vTpmDt3LuLi4jBv3jwEBgYiIiIiz1jT09ORkpKisBAREVHlJRMKe0dD5UpqaiqMjY2xYcMG9O7dGwDw/PlzVK9eHcOHD8eyZctgY2MDf39/+Pv747vvvsOaNWtw7do1qKqq5mjv47r58fb2hr29PRYvXgzgw7fymZmZOHXqlFinSZMmaNOmDUJDQ3Ho0CF06tQJcXFxsLOzy9Gej48PNDU1sWbNGrHs9OnTcHd3R1paGjQ0NPKNx9/fH7GxsYiMjFToS8OGDbFz506xrH///njy5AkOHToklgUEBODPP//E9evXxf1atmyJjRs3AgAePXqEqlWrIjAwECEhIQCAc+fOiYmHubl5vrGFh4dj8ODBuHPnDmrVqgUAWLVqFUJCQvDo0SMAgIWFBcaMGYP//e9/CuevcePG+OGHH5CYmIgaNWpg3bp1GDp0KABgy5Yt6Nu3L44ePYo2bdoAAEJDQxEeHo6///4bAGBlZYUFCxagb9++Yrtz5szBvn37EBUVlWu8QUFBCA4OzlFu6b8NSupa+faVKrbEUO+yDoGIiEpISkoK9PX1kZycDD09vXzr8olABRUfH493797Bzc1NLDMyMkKdOnVyrd+zZ0+8efMGNWvWxLBhw7Bz505kZGTke4zMzEzMnTsXzs7OMDY2ho6ODg4dOoT79+8r1HN2dlZYr1q1Kh4/fgwAiI2NRfXq1XNNAgAgJiYG4eHh0NHRERdPT0/I5XIkJCQUeB7y4urqqrAeFxeHFi1aKJS1aNECt2/fRmZmZq59MTMzAwDxycnHZVn9K4iWlpaYBACK5yYlJQX//vtvrnHFxcUplBUmrqx2nzx5ggcPHmDo0KEK53XOnDmIj4/PM9Zp06YhOTlZXB48eFCoPhIREVHFxJeFK6iiPsixtLTEzZs3cfjwYRw5cgSjR4/GokWLcOLEiVyfEADAkiVLsHTpUixbtgxOTk7Q1taGv78/3r17p1Av+/4ymQxyuRwAoKmpmW9ccrkcI0aMEN9X+JiVlVVRuqhAW1tbYV0QBPHdh4/Lsvu4L1n1cyvL6l9Bcjs32Y+bW1zZywoTV1ZMWf9du3YtmjZtqtBO1pCs3Kirq0NdXT3f/hAREVHlwUSggrK1tYWqqirOnTsn3jC/ePECt27dgru7e677aGpqomvXrujatSvGjBmDunXr4urVq3BxcYGamprCN+MAcOrUKXTr1g0DBgwA8OEG8/bt27C3ty90nM7Ozvjnn39w69atXJ8KuLi44Pr167C1tS10m8Xh4OCA06dPK5RFRUXBzs4u35vj0qSnp4dq1arh9OnTaNWqlUJcTZo0KXa7ZmZmsLCwwN27d9G/f/+SCJWIiIgqISYCFZSOjg6GDh2KyZMnw9jYGGZmZpg+fTqUlHIf7RUeHo7MzEw0bdoUWlpa2LhxIzQ1NWFtbQ3gw/j4kydPok+fPlBXV0eVKlVga2uL7du3IyoqCoaGhvjuu+/w6NGjIiUC7u7uaNWqFXr06IHvvvsOtra2+PvvvyGTydCxY0dMmTIFzZo1w5gxYzBs2DBoa2sjLi4Ohw8fxsqVK0vkXAHAxIkT0bhxY8yePRu9e/fG2bNn8f3335f5TDqTJ0/GrFmzUKtWLTRo0ABhYWGIjY3Fpk2bPqndoKAg+Pn5QU9PD506dUJ6ejqio6Px4sULTJgwoYSiJyIiooqMiUAFtmjRIqSmpqJr167Q1dXFxIkTkZycnGtdAwMDhIaGYsKECcjMzISTkxP27NkDY2NjAEBISAhGjBiBWrVqIT09HYIgIDAwEAkJCfD09ISWlhaGDx+O7t2753mMvGzfvh2TJk1C3759kZaWBltbW4SGhgL48MTgxIkTmD59Olq2bAlBEFCrVi3xBeiS4uLigm3btmHmzJmYPXs2qlatipCQEPj6+pbocYrKz88PKSkpmDhxIh4/fgwHBwfs3r0btWvX/qR2v/nmG2hpaWHRokUICAiAtrY2nJycCnwZnIiIiKSDswYRUa6yZh3grEGVH2cNIiKqPDhrEBERERER5YuJAJVbp06dUpj+MvtS1jp16pRnbPPmzSvr8IiIiIjyxXcEqNxydXVFbGxsWYeRp3Xr1uHNmze5bjMyMvrM0RAREREVDRMBKrc0NTVLfVrRT2FhYVHWIRAREREVG4cGERERERFJEBMBIiIiIiIJYiJARERERCRBTASIiIiIiCSIiQARERERkQQxESAiIiIikiAmAkREREREEsREgIiIiIhIgpgIEBERERFJEH9ZmIjydS3YE3p6emUdBhEREZUwPhEgIiIiIpIgJgJERERERBLERICIiIiISIKYCBARERERSRATASIiIiIiCWIiQEREREQkQUwEiIiIiIgkiIkAEREREZEEMREgIiIiIpIg/rIwEeXLcdZBKKlrlXUYRCQhiaHeZR0CkSTwiQARERERkQQxESAiIiIikiAmAkREREREEsREgIiIiIhIgpgIEBERERFJEBMBIiIiIiIJYiJARERERCRBTASIiIiIiCSIiQARERERkQQxESAiIiIikiAmAkREREREEsREgIiIiIhIgpgIEBERERFJEBMBIiIiIiIJYiJAkhAUFIQGDRqI676+vujevXuZxfMpsveFiIiIqDhUyjoAorKwfPlyCIJQ1mEQERERlRkmAiRJ+vr6ZR1Cgd6/fw9VVdUK1zYRERFVDBwaRBVCeno6/Pz8YGpqCg0NDXzxxRe4cOECACAyMhIymQxHjx6Fq6srtLS00Lx5c9y8eTPP9rIPDfLw8ICfnx8CAgJgZGQEc3NzBAUFKeyTnJyM4cOHw9TUFHp6emjTpg0uX75c6D6sXr0atWrVgpqaGurUqYONGzcqbJfJZPjxxx/RrVs3aGtrY86cOQCA0NBQmJmZQVdXF0OHDsXbt29ztB0WFgZ7e3toaGigbt26WLVqlbgtMTERMpkM27Ztg4eHBzQ0NPDLL78UOm4iIiKqnJgIUIUQEBCA7du3IyIiAhcvXoStrS08PT3x/Plzsc706dOxZMkSREdHQ0VFBUOGDCnSMSIiIqCtrY3z589j4cKFCAkJweHDhwEAgiDA29sbjx49wr59+xATEwMXFxe0bdtWIYa87Ny5E+PGjcPEiRNx7do1jBgxAoMHD8bx48cV6s2aNQvdunXD1atXMWTIEGzbtg2zZs3C3LlzER0djapVqyrc5APA2rVrMX36dMydOxdxcXGYN28eAgMDERERoVBvypQp8PPzQ1xcHDw9PXPEmJ6ejpSUFIWFiIiIKi+ZwIHSVM6lpaXB0NAQ4eHh6NevH4APQ1tsbGzg7++Pxo0bo3Xr1jhy5Ajatm0LANi3bx+8vb3x5s0baGhoICgoCLt27UJsbCyAD08EXr58iV27dgH48EQgMzMTp06dEo/bpEkTtGnTBqGhoTh27Bi+/PJLPH78GOrq6mIdW1tbBAQEYPjw4fn2oUWLFqhXrx5++uknsaxXr15IS0vDn3/+CeDDEwF/f38sXbpUrNO8eXPUr18fq1evFsuaNWuGt2/fin2xsrLCggUL0LdvX7HOnDlzsG/fPkRFRSExMRE1atTAsmXLMG7cuDxjDAoKQnBwcI5yS/9tUFLXyrd/REQlKTHUu6xDIKqwUlJSoK+vj+TkZOjp6eVbl08EqNyLj4/H+/fv0aJFC7FMVVUVTZo0QVxcnFjm7Ows/rlq1aoAgMePHxf6OB/vn9VG1v4xMTFITU2FsbExdHR0xCUhIQHx8fEFth0XF6cQP/AhOfg4fgBwdXXNsZ+bm5tC2cfrT548wYMHDzB06FCFuObMmZMjruxtZzdt2jQkJyeLy4MHDwrsFxEREVVcfFmYyr2sh1YymSxH+cdlH7/8mlUul8sLfZzsL8/KZDJxf7lcjqpVqyIyMjLHfgYGBoVqv6D4AUBbW7vQ8WbFBXwYHtS0aVOFbcrKykVqW11dXeFpBxEREVVufCJA5Z6trS3U1NRw+vRpsez9+/eIjo6Gvb39Z4nBxcUFjx49goqKCmxtbRWWKlWqFLi/vb29QvwAEBUVVWD89vb2OHfunELZx+tmZmawsLDA3bt3c8RVo0aNIvSQiIiIpIZPBKjc09bWxqhRozB58mQYGRnBysoKCxcuxOvXrzF06NAizdxTXO3atYObmxu6d++OBQsWoE6dOvj333+xb98+dO/evcBhN5MnT0avXr3EF4z37NmDHTt24MiRI/nuN27cOAwaNAiurq744osvsGnTJly/fh01a9YU6wQFBcHPzw96enro1KkT0tPTER0djRcvXmDChAkl0n8iIiKqfJgIUIUQGhoKuVyOgQMH4tWrV3B1dcXBgwdhaGj4WY4vk8mwb98+TJ8+HUOGDMGTJ09gbm6OVq1awczMrMD9u3fvjuXLl2PRokXw8/NDjRo1EBYWBg8Pj3z36927N+Lj4zFlyhS8ffsWPXr0wKhRo3Dw4EGxzjfffAMtLS0sWrQIAQEB0NbWhpOTE/z9/T+x10RERFSZcdYgIspV1qwDnDWIiD43zhpEVHycNYiIiIiIiPLFRICoBNSrV09h+s6Pl02bNpV1eEREREQ58B0BohKwb98+vH//PtdthXmHgIiIiOhzYyJAVAKsra3LOgQiIiKiIuHQICIiIiIiCWIiQEREREQkQUwEiIiIiIgkiIkAEREREZEEMREgIiIiIpIgJgJERERERBLERICIiIiISIKYCBARERERSRATASIiIiIiCeIvCxNRvq4Fe0JPT6+swyAiIqISxicCREREREQSxESAiIiIiEiCmAgQEREREUkQEwEiIiIiIgliIkBEREREJEFMBIiIiIiIJIiJABERERGRBDERICIiIiKSIP6gGBHly3HWQSipa5V1GERERJVGYqh3WYcAgE8EiIiIiIgkiYkAEREREZEEMREgIiIiIpIgJgJERERERBLERICIiIiISIKYCBARERERSRATASIiIiIiCWIiQEREREQkQUwEiIiIiIgkiIkAEREREZEEMREgIiIiIpIgJgJERERERBLERICIiIiISIKYCBARERERSVClSQRsbGywbNmysg6jXAgKCkKDBg3EdV9fX3Tv3r3M4vmcKnNfZTIZdu3aVdZhEBERUSWhUtYBFFV4eDj8/f3x8uVLhfILFy5AW1u7RI4RFBSEXbt2ITY2tkTaK2vLly+HIAhlHQZ9oqSkJBgaGpZ1GERERFRJlKtE4N27d1BTUyvWviYmJiUcTeWhr69f1iGUiE/5fFRkWf02Nzcv61CIiIioEinToUEeHh4YO3YsJkyYgCpVqqB9+/b47rvv4OTkBG1tbVhaWmL06NFITU0FAERGRmLw4MFITk6GTCaDTCZDUFAQgJxDg+7fv49u3bpBR0cHenp66NWrF/77778CYwoPD0dwcDAuX74sHiM8PPyT2gT+b7jO+vXrYWVlBR0dHYwaNQqZmZlYuHAhzM3NYWpqirlz5yrsl5ycjOHDh8PU1BR6enpo06YNLl++rFAnNDQUZmZm0NXVxdChQ/H27VuF7dmHyxw4cABffPEFDAwMYGxsjM6dOyM+Pl7cnpiYCJlMhh07dqB169bQ0tJC/fr1cfbs2UL1NTw8HAYGBti1axfs7OygoaGB9u3b48GDB3nGBAD+/v7w8PAQ13P7fADA9evX4e3tDT09Pejq6qJly5YK8QPA4sWLUbVqVRgbG2PMmDF4//69uO2XX36Bq6srdHV1YW5ujn79+uHx48fi9hcvXqB///4wMTGBpqYmateujbCwMHH7w4cP0bt3bxgaGsLY2BjdunVDYmJioc5NVr+Dg4PFazpixAi8e/euwH5nHxr0zz//oE+fPjAyMoK2tjZcXV1x/vx5cfuePXvQqFEjaGhooGbNmggODkZGRkah4iQiIqLKr8zfEYiIiICKigrOnDmDNWvWQElJCStWrMC1a9cQERGBY8eOISAgAADQvHlzLFu2DHp6ekhKSkJSUhImTZqUo01BENC9e3c8f/4cJ06cwOHDhxEfH4/evXsXGE/v3r0xceJE1KtXTzxG7969P6nNLPHx8di/fz8OHDiAzZs3Y/369fD29sY///yDEydOYMGCBZgxYwbOnTsn9sPb2xuPHj3Cvn37EBMTAxcXF7Rt2xbPnz8HAGzbtg2zZs3C3LlzER0djapVq2LVqlX5xpGWloYJEybgwoULOHr0KJSUlPDll19CLpcr1Js+fTomTZqE2NhY2NnZoW/fvoW+kXz9+jXmzp2LiIgInDlzBikpKejTp0+hz1WW7J+Phw8folWrVtDQ0MCxY8cQExODIUOGKMR1/PhxxMfH4/jx44iIiEB4eLiYzAEfvmGfPXs2Ll++jF27diEhIQG+vr7i9sDAQNy4cQP79+9HXFwcVq9ejSpVqoj9at26NXR0dHDy5EmcPn0aOjo66Nixo8LNfH6OHj2KuLg4HD9+HJs3b8bOnTsRHBycb7+zS01Nhbu7O/7991/s3r0bly9fRkBAgHgNDx48iAEDBsDPzw83btzAmjVrEB4eniPR/Fh6ejpSUlIUFiIiIqq8ynxokK2tLRYuXCiu161bV/xzjRo1MHv2bIwaNQqrVq2Cmpoa9PX1IZPJ8h0mceTIEVy5cgUJCQmwtLQEAGzcuBH16tXDhQsX0Lhx4zz31dTUhI6ODlRUVBSOcfjw4WK3mUUul2P9+vXQ1dWFg4MDWrdujZs3b2Lfvn1QUlJCnTp1sGDBAkRGRqJZs2Y4fvw4rl69isePH0NdXR3Ah2+6d+3ahd9//x3Dhw/HsmXLMGTIEHzzzTcAgDlz5uDIkSM5ngp8rEePHgrrP//8M0xNTXHjxg04OjqK5ZMmTYK3tzcAIDg4GPXq1cOdO3cUrlFe3r9/j++//x5NmzYF8OHG1t7eHn/99ReaNGlS4P5Zsn8+/ve//0FfXx9btmyBqqoqAMDOzk5hH0NDQ3z//fdQVlZG3bp14e3tjaNHj2LYsGEAgCFDhoh1a9asiRUrVqBJkyZITU2Fjo4O7t+/j4YNG8LV1RXAh6dNWbZs2QIlJSWsW7cOMpkMABAWFgYDAwNERkaiQ4cOBfZJTU0N69evh5aWFurVq4eQkBBMnjwZs2fPhpKSUq79zu7XX3/FkydPcOHCBRgZGYn7ZJk7dy6mTp2KQYMGif2cPXs2AgICMGvWrFzbnD9/fo6EhIiIiCqvMn8ikHWzleX48eNo3749LCwsoKurCx8fHzx79gxpaWmFbjMuLg6WlpbiDTsAODg4wMDAAHFxccWKsyTatLGxga6urrhuZmYGBwcH8eYvqyxrmEpMTAxSU1NhbGwMHR0dcUlISBCHwsTFxcHNzU3hONnXs4uPj0e/fv1Qs2ZN6OnpoUaNGgA+DH36mLOzs/jnqlWrAoDCEJr8qKioKFzbunXrFuv8Z/98xMbGomXLlmISkJt69epBWVlZXK9atapC3JcuXUK3bt1gbW0NXV1dcThSVv9HjRqFLVu2oEGDBggICEBUVJS4b0xMDO7cuQNdXV3xehgZGeHt27c5hiflpX79+tDS0hLX3dzckJqaqjB0Knu/s4uNjUXDhg3FJCC7mJgYhISEKHxuhg0bhqSkJLx+/TrXfaZNm4bk5GRx+TgeIiIiqnzK/InAxzP93Lt3D15eXhg5ciRmz54NIyMjnD59GkOHDlUY410QQRDEb2sLU/652sx+8yqTyXItyxreIZfLUbVqVURGRuZoy8DAoHCB56JLly6wtLTE2rVrUa1aNcjlcjg6OuYY2vJxbFl9zD58KD+5nZesMiUlpRwzGeV2jbPPBKWpqVngcfM7p2lpaejQoQM6dOiAX375BSYmJrh//z48PT3F/nfq1An37t3Dn3/+iSNHjqBt27YYM2YMFi9eDLlcjkaNGmHTpk05jvupL6x/fL4KmgGroPMgl8sRHByMr776Ksc2DQ2NXPdRV1cXnzwRERFR5VfmicDHoqOjkZGRgSVLlojfkm/btk2hjpqaGjIzM/Ntx8HBAffv38eDBw/Eb/Bv3LiB5ORk2NvbFxhHbsf41DaLw8XFBY8ePYKKiorC8JSP2dvb49y5c/Dx8RHLst4xyM2zZ88QFxeHNWvWoGXLlgCA06dPl2jcAJCRkYHo6GhxGNDNmzfx8uVLcViRiYkJrl27prBPbGxsvt/0Ax+eUkREROD9+/cF1s3N33//jadPnyI0NFS8jtHR0TnqmZiYwNfXF76+vmjZsiUmT56MxYsXw8XFBVu3bhVf9C2Oy5cv482bN+LN/Llz56Cjo4Pq1asXug1nZ2esW7cOz58/z/WpgIuLC27evKkwXIiIiIjoY2U+NOhjtWrVQkZGBlauXIm7d+9i48aN+PHHHxXq2NjYIDU1FUePHsXTp09zHebQrl07ODs7o3///rh48SL++usv+Pj4wN3dvcAhF1nHSEhIQGxsLJ4+fYr09PRPbrM42rVrBzc3N3Tv3h0HDx5EYmIioqKiMGPGDPHmddy4cVi/fj3Wr1+PW7duYdasWbh+/XqebWbNdPPTTz/hzp07OHbsGCZMmFDisauqquLbb7/F+fPncfHiRQwePBjNmjUTE4M2bdogOjoaGzZswO3btzFr1qwciUFuxo4dK754HB0djdu3b2Pjxo24efNmoeKysrKCmpqa+BnbvXs3Zs+erVBn5syZ+OOPP3Dnzh1cv34de/fuFZO9/v37o0qVKujWrRtOnTqFhIQEnDhxAuPGjcM///xTqBjevXuHoUOHii8kz5o1C2PHjlUYIlaQvn37wtzcHN27d8eZM2dw9+5dbN++XZzZaebMmdiwYQOCgoJw/fp1xMXFYevWrZgxY0ahj0FERESVW7lKBBo0aIDvvvsOCxYsgKOjIzZt2oT58+cr1GnevDlGjhyJ3r17w8TEJNcXKrOmWTQ0NESrVq3Qrl071KxZE1u3bi1UHD169EDHjh3RunVrmJiYYPPmzZ/cZnHIZDLs27cPrVq1wpAhQ2BnZ4c+ffogMTERZmZmAD7McjRz5kxMmTIFjRo1wr179zBq1Kg821RSUsKWLVsQExMDR0dHjB8/HosWLSrx2LW0tDBlyhT069cPbm5u0NTUxJYtW8Ttnp6eCAwMREBAABo3boxXr14pPNXIi7GxMY4dOybOmtOoUSOsXbu20E8HTExMEB4ejt9++w0ODg4IDQ3F4sWLFeqoqalh2rRpcHZ2RqtWraCsrCzGrqWlhZMnT8LKygpfffUV7O3tMWTIELx586bQTwjatm2L2rVro1WrVujVqxe6dOkiToNbWGpqajh06BBMTU3h5eUFJycnhIaGiu9GeHp6Yu/evTh8+DAaN26MZs2a4bvvvoO1tXWRjkNERESVl0zgT85SCcvr15/pw+8IvHz5UuH3AMqrlJQU6Ovrw9J/G5TUtQregYiIiAolMdS71NrO+vc7OTm5wC8py9UTASIiIiIi+jwkmQjUq1dPYVrFj5fcZoMpqzbLq06dOuXZ13nz5pV1eGUqr/Oio6ODU6dOlXV4RERERCJJDg26d+9entORmpmZKcz1X5ZtllcPHz7Emzdvct1mZGSU59z2UnDnzp08t1lYWBRq+tPygkODiIiISkd5GRpUrqYP/VxK44VJKb2EaWFhUdYhlFucrpOIiIgqCkkODSIiIiIikjomAkREREREEsREgIiIiIhIgpgIEBERERFJEBMBIiIiIiIJYiJARERERCRBTASIiIiIiCSIiQARERERkQQxESAiIiIikiBJ/rIwERXetWDPAn+inIiIiCoePhEgIiIiIpIgJgJERERERBLERICIiIiISIKYCBARERERSRATASIiIiIiCWIiQEREREQkQUwEiIiIiIgkiIkAEREREZEEMREgIiIiIpIgJgJERERERBLERICIiIiISIKYCBARERERSRATASIiIiIiCWIiQEREREQkQUwEiIiIiIgkSKWsAyCi8kkQBABASkpKGUdCREREhZX173bWv+P5YSJARLl69uwZAMDS0rKMIyEiIqKievXqFfT19fOtw0SAiHJlZGQEALh//36B/yOpiFJSUmBpaYkHDx5AT0+vrMMpcexfxVWZ+wawfxUd+1f+CYKAV69eoVq1agXWZSJARLlSUvrwCpG+vn6F/Z9hYejp6bF/FVhl7l9l7hvA/lV07F/5Vtgv8PiyMBERERGRBDERICIiIiKSICYCRJQrdXV1zJo1C+rq6mUdSqlg/yq2yty/ytw3gP2r6Ni/ykUmFGZuISIiIiIiqlT4RICIiIiISIKYCBARERERSRATASIiIiIiCWIiQEREREQkQUwEiCRi1apVqFGjBjQ0NNCoUSOcOnUq3/onTpxAo0aNoKGhgZo1a+LHH3/MUWf79u1wcHCAuro6HBwcsHPnztIKv0Al3b/w8HDIZLIcy9u3b0uzG3kqSv+SkpLQr18/1KlTB0pKSvD398+1XkW9foXpX0W+fjt27ED79u1hYmICPT09uLm54eDBgznqVdTrV5j+lafrV5S+nT59Gi1atICxsTE0NTVRt25dLF26NEe9inrtCtO/8nTtgKL/25DlzJkzUFFRQYMGDXJsK0/X75MJRFTpbdmyRVBVVRXWrl0r3LhxQxg3bpygra0t3Lt3L9f6d+/eFbS0tIRx48YJN27cENauXSuoqqoKv//+u1gnKipKUFZWFubNmyfExcUJ8+bNE1RUVIRz5859rm6JSqN/YWFhgp6enpCUlKSwlIWi9i8hIUHw8/MTIiIihAYNGgjjxo3LUaciX7/C9K8iX79x48YJCxYsEP766y/h1q1bwrRp0wRVVVXh4sWLYp2KfP0K07/ycv2K2reLFy8Kv/76q3Dt2jUhISFB2Lhxo6ClpSWsWbNGrFORr11h+lderp0gFL1/WV6+fCnUrFlT6NChg1C/fn2FbeXp+pUEJgJEEtCkSRNh5MiRCmV169YVpk6dmmv9gIAAoW7dugplI0aMEJo1ayau9+rVS+jYsaNCHU9PT6FPnz4lFHXhlUb/wsLCBH19/RKPtTiK2r+Pubu753qjXJGv38fy6l9luX5ZHBwchODgYHG9sly/LNn7V16uX0n07csvvxQGDBggrle2a5e9f+Xl2glC8fvXu3dvYcaMGcKsWbNyJALl6fqVBA4NIqrk3r17h5iYGHTo0EGhvEOHDoiKisp1n7Nnz+ao7+npiejoaLx//z7fOnm1WVpKq38AkJqaCmtra1SvXh2dO3fGpUuXSr4DBShO/wqjIl+/wqos108ul+PVq1cwMjISyyrT9cutf0DZX7+S6NulS5cQFRUFd3d3sawyXbvc+geU/bUDit+/sLAwxMfHY9asWbluLy/Xr6QwESCq5J4+fYrMzEyYmZkplJuZmeHRo0e57vPo0aNc62dkZODp06f51smrzdJSWv2rW7cuwsPDsXv3bmzevBkaGhpo0aIFbt++XTodyUNx+lcYFfn6FUZlun5LlixBWloaevXqJZZVpuuXW//Kw/X7lL5Vr14d6urqcHV1xZgxY/DNN9+I2yrDtcuvf+Xh2gHF69/t27cxdepUbNq0CSoqKrnWKS/Xr6Tk3ksiqnRkMpnCuiAIOcoKqp+9vKhtlqaS7l+zZs3QrFkzcXuLFi3g4uKClStXYsWKFSUVdqGVxrmuyNevIJXl+m3evBlBQUH4448/YGpqWiJtloaS7l95un7F6dupU6eQmpqKc+fOYerUqbC1tUXfvn0/qc3SUtL9K0/XDih8/zIzM9GvXz8EBwfDzs6uRNqsCJgIEFVyVapUgbKyco5vKx4/fpzjW40s5ubmudZXUVGBsbFxvnXyarO0lFb/slNSUkLjxo0/+7daxelfYVTk61ccFfH6bd26FUOHDsVvv/2Gdu3aKWyrDNcvv/5lVxbX71P6VqNGDQCAk5MT/vvvPwQFBYk3ypXh2uXXv+wqyt+9V69eITo6GpcuXcLYsWMBfBi2JggCVFRUcOjQIbRp06bcXL+SwqFBRJWcmpoaGjVqhMOHDyuUHz58GM2bN891Hzc3txz1Dx06BFdXV6iqquZbJ682S0tp9S87QRAQGxuLqlWrlkzghVSc/hVGRb5+xVHRrt/mzZvh6+uLX3/9Fd7e3jm2V/TrV1D/siuL61dSn01BEJCeni6uV/Rrl132/uW2vSL83dPT08PVq1cRGxsrLiNHjkSdOnUQGxuLpk2bAig/16/EfMYXk4mojGRNofbzzz8LN27cEPz9/QVtbW0hMTFREARBmDp1qjBw4ECxftb0muPHjxdu3Lgh/Pzzzzmm1zxz5oygrKwshIaGCnFxcUJoaGiZT4FXkv0LCgoSDhw4IMTHxwuXLl0SBg8eLKioqAjnz58v9/0TBEG4dOmScOnSJaFRo0ZCv379hEuXLgnXr18Xt1fk6ycIBfevIl+/X3/9VVBRURF++OEHhekXX758KdapyNevMP0rL9evqH37/vvvhd27dwu3bt0Sbt26Jaxfv17Q09MTpk+fLtapyNeuMP0rL9euOP3LLrdZg8rT9SsJTASIJOKHH34QrK2tBTU1NcHFxUU4ceKEuG3QoEGCu7u7Qv3IyEihYcOGgpqammBjYyOsXr06R5u//fabUKdOHUFVVVWoW7eusH379tLuRp5Kun/+/v6ClZWVoKamJpiYmAgdOnQQoqKiPkdXclXU/gHIsVhbWyvUqcjXr6D+VeTr5+7unmv/Bg0apNBmRb1+helfebp+RenbihUrhHr16glaWlqCnp6e0LBhQ2HVqlVCZmamQpsV9doVpn/l6doJQtH/3/Kx3BIBQShf1+9TyQTh/78hR0REREREksF3BIiIiIiIJIiJABERERGRBDERICIiIiKSICYCREREREQSxESAiIiIiEiCmAgQEREREUkQEwEiIiIiIgliIkBEREREJEFMBIiIiEqRr68vunfvXtZh5CoxMREymQyxsbFlHQoRlQEmAkRERBL07t27sg6BiMoYEwEiIqLPxMPDA99++y38/f1haGgIMzMz/PTTT0hLS8PgwYOhq6uLWrVqYf/+/eI+kZGRkMlk+PPPP1G/fn1oaGigadOmuHr1qkLb27dvR7169aCurg4bGxssWbJEYbuNjQ3mzJkDX19f6OvrY9iwYahRowYAoGHDhpDJZPDw8AAAXLhwAe3bt0eVKlWgr68Pd3d3XLx4UaE9mUyGdevW4csvv4SWlhZq166N3bt3K9S5fv06vL29oaenB11dXbRs2RLx8fHi9rCwMNjb20NDQwN169bFqlWrPvkcE1HhMREgIiL6jCIiIlClShX89ddf+PbbbzFq1Cj07NkTzZs3x8WLF+Hp6YmBAwfi9evXCvtNnjwZixcvxoULF2BqaoquXbvi/fv3AICYmBj06tULffr0wdWrVxEUFITAwECEh4crtLFo0SI4OjoiJiYGgYGB+OuvvwAAR44cQVJSEnbs2AEAePXqFQYNGoRTp07h3LlzqF27Nry8vPDq1SuF9oKDg9GrVy9cuXIFXl5e6N+/P54/fw4AePjwIVq1agUNDQ0cO3YMMTExGDJkCDIyMgAAa9euxfTp0zF37lzExcVh3rx5CAwMRERERImfcyLKg0BERESlZtCgQUK3bt0EQRAEd3d34YsvvhC3ZWRkCNra2sLAgQPFsqSkJAGAcPbsWUEQBOH48eMCAGHLli1inWfPngmamprC1q1bBUEQhH79+gnt27dXOO7kyZMFBwcHcd3a2lro3r27Qp2EhAQBgHDp0qV8+5CRkSHo6uoKe/bsEcsACDNmzBDXU1NTBZlMJuzfv18QBEGYNm2aUKNGDeHdu3e5tmlpaSn8+uuvCmWzZ88W3Nzc8o2FiEoOnwgQERF9Rs7OzuKflZWVYWxsDCcnJ7HMzMwMAPD48WOF/dzc3MQ/GxkZoU6dOoiLiwMAxMXFoUWLFgr1W7Rogdu3byMzM1Msc3V1LVSMjx8/xsiRI2FnZwd9fX3o6+sjNTUV9+/fz7Mv2tra0NXVFeOOjY1Fy5YtoaqqmqP9J0+e4MGDBxg6dCh0dHTEZc6cOQpDh4iodKmUdQBERERSkv3GWCaTKZTJZDIAgFwuL7CtrLqCIIh/ziIIQo762trahYrR19cXT548wbJly2BtbQ11dXW4ubnleME4t75kxa2pqZln+1l11q5di6ZNmypsU1ZWLlSMRPTpmAgQERFVAOfOnYOVlRUA4MWLF7h16xbq1q0LAHBwcMDp06cV6kdFRcHOzi7fG2s1NTUAUHhqAACnTp3CqlWr4OXlBQB48OABnj59WqR4nZ2dERERgffv3+dIGMzMzGBhYYG7d++if//+RWqXiEoOEwEiIqIKICQkBMbGxjAzM8P06dNRpUoV8fcJJk6ciMaNG2P27Nno3bs3zp49i++//77AWXhMTU2hqamJAwcOoHr16tDQ0IC+vj5sbW2xceNGuLq6IiUlBZMnT873G/7cjB07FitXrkSfPn0wbdo06Ovr49y5c2jSpAnq1KmDoKAg+Pn5QU9PD506dUJ6ejqio6Px4sULTJgwobiniYiKgO8IEBERVQChoaEYN24cGjVqhKSkJOzevVv8Rt/FxQXbtm3Dli1b4OjoiJkzZyIkJAS+vr75tqmiooIVK1ZgzZo1qFatGrp16wYAWL9+PV68eIGGDRti4MCB8PPzg6mpaZHiNTY2xrFjx5Camgp3d3c0atQIa9euFZ8OfPPNN1i3bh3Cw8Ph5OQEd3d3hIeHi1OaElHpkwm5DSIkIiKiciEyMhKtW7fGixcvYGBgUNbhEFElwicCREREREQSxESAiIiIiEiCODSIiIiIiEiC+ESAiIiIiEiCmAgQEREREUkQEwEiIiIiIgliIkBEREREJEFMBIiIiIiIJIiJABERERGRBDERICIiIiKSICYCREREREQSxESAiIiIiEiC/h9jvZ8771RgmQAAAABJRU5ErkJggg==", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Create a DataFrame with the feature importance\n", "df_feature_importance_dt = pd.DataFrame({'Feature': X_train.columns, 'Importance': clf_dt.feature_importances_})\n", "df_feature_importance_dt = df_feature_importance_dt.sort_values('Importance', ascending=False)\n", "\n", "# Plot the feature importance\n", "plt.barh(df_feature_importance_dt['Feature'], df_feature_importance_dt['Importance'])\n", "plt.xlabel('Importance')\n", "plt.ylabel('Feature')\n", "plt.title('Feature Importance - Decision Tree')\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This shows that the `ratio_to_median_purchase_price` is the most important feature for determining whether a transaction is fraudulent or not. Whether a transaction is online, is important as well.\n", "\n", "Let's also look at the feature importance of the Random Forest model" ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:43:37.561504Z", "iopub.status.busy": "2024-05-31T21:43:37.561244Z", "iopub.status.idle": "2024-05-31T21:43:37.704444Z", "shell.execute_reply": "2024-05-31T21:43:37.703654Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Create a DataFrame with the feature importance\n", "df_feature_importance_rf = pd.DataFrame({'Feature': X_train.columns, 'Importance': clf_rf.feature_importances_})\n", "df_feature_importance_rf = df_feature_importance_rf.sort_values('Importance', ascending=False)\n", "\n", "# Plot the feature importance\n", "plt.barh(df_feature_importance_rf['Feature'], df_feature_importance_rf['Importance'])\n", "plt.xlabel('Importance')\n", "plt.ylabel('Feature')\n", "plt.title('Feature Importance - Random Forest')\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Somewhat surprisingly, XGBoost seems to have picked up on different features than the Decision Tree and Random Forest models. The most important feature is `online_order`, followed by `ratio_to_median_purchase_price` as you can see from the plot below" ] }, { "cell_type": "code", "execution_count": 21, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:43:37.707918Z", "iopub.status.busy": "2024-05-31T21:43:37.707634Z", "iopub.status.idle": "2024-05-31T21:43:37.851716Z", "shell.execute_reply": "2024-05-31T21:43:37.850888Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Create a DataFrame with the feature importance\n", "df_feature_importance_xgb = pd.DataFrame({'Feature': X_train.columns, 'Importance': clf_xgb.feature_importances_})\n", "df_feature_importance_xgb = df_feature_importance_xgb.sort_values('Importance', ascending=False)\n", "\n", "# Plot the feature importance\n", "plt.barh(df_feature_importance_xgb['Feature'], df_feature_importance_xgb['Importance'])\n", "plt.xlabel('Importance')\n", "plt.ylabel('Feature')\n", "plt.title('Feature Importance - XGBoost')\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Permuation Importance\n", "\n", "We can also look at the permutation importance of each model. The permutation importance is a measure of how much each feature contributes to the model's predictions. The permutation importance is calculated by permuting the values of each feature and measuring how much the model's performance decreases. Let's start with the Decision Tree model" ] }, { "cell_type": "code", "execution_count": 22, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:43:37.856684Z", "iopub.status.busy": "2024-05-31T21:43:37.856441Z", "iopub.status.idle": "2024-05-31T21:43:42.856282Z", "shell.execute_reply": "2024-05-31T21:43:42.855360Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Calculate the permutation importance\n", "result_dt = permutation_importance(clf_dt, X_test, y_test, n_repeats=10, random_state=0)\n", "\n", "# Create a DataFrame with the permutation importance\n", "df_permutation_importance_dt = pd.DataFrame({'Feature': X_train.columns, 'Importance': result_dt.importances_mean})\n", "df_permutation_importance_dt = df_permutation_importance_dt.sort_values('Importance', ascending=False)\n", "\n", "# Plot the permutation importance\n", "plt.barh(df_permutation_importance_dt['Feature'], df_permutation_importance_dt['Importance'])\n", "plt.xlabel('Accuracy Decrease')\n", "plt.ylabel('Feature')\n", "plt.title('Permutation Importance - Decision Tree')\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's also look at the permutation importance of the Random Forest model" ] }, { "cell_type": "code", "execution_count": 23, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:43:42.859962Z", "iopub.status.busy": "2024-05-31T21:43:42.859708Z", "iopub.status.idle": "2024-05-31T21:45:30.733357Z", "shell.execute_reply": "2024-05-31T21:45:30.732541Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Calculate the permutation importance\n", "result_rf = permutation_importance(clf_rf, X_test, y_test, n_repeats=10, random_state=0)\n", "\n", "# Create a DataFrame with the permutation importance\n", "df_permutation_importance_rf = pd.DataFrame({'Feature': X_train.columns, 'Importance': result_rf.importances_mean})\n", "df_permutation_importance_rf = df_permutation_importance_rf.sort_values('Importance', ascending=False)\n", "\n", "# Plot the permutation importance\n", "plt.barh(df_permutation_importance_rf['Feature'], df_permutation_importance_rf['Importance'])\n", "plt.xlabel('Accuracy Decrease')\n", "plt.ylabel('Feature')\n", "plt.title('Permutation Importance - Random Forest')\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's also look at the permutation importance of the XGBoost model" ] }, { "cell_type": "code", "execution_count": 24, "metadata": { "execution": { "iopub.execute_input": "2024-05-31T21:45:30.738393Z", "iopub.status.busy": "2024-05-31T21:45:30.738183Z", "iopub.status.idle": "2024-05-31T21:45:39.521403Z", "shell.execute_reply": "2024-05-31T21:45:39.520719Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Calculate the permutation importance\n", "result_xgb = permutation_importance(clf_xgb, X_test, y_test, n_repeats=10, random_state=0)\n", "\n", "# Create a DataFrame with the permutation importance\n", "df_permutation_importance_xgb = pd.DataFrame({'Feature': X_train.columns, 'Importance': result_xgb.importances_mean})\n", "df_permutation_importance_xgb = df_permutation_importance_xgb.sort_values('Importance', ascending=False)\n", "\n", "# Plot the permutation importance\n", "plt.barh(df_permutation_importance_xgb['Feature'], df_permutation_importance_xgb['Importance'])\n", "plt.xlabel('Accuracy Decrease')\n", "plt.ylabel('Feature')\n", "plt.title('Permutation Importance - XGBoost')\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Here the results for the three models are quite similar. The most important feature is `ratio_to_median_purchase_price`, followed by `online_order`.\n", "\n", "\n", "### Conclusions\n", "\n", "In this notebook, we have seen how to implement decision trees, random forests, and XGBoost classifiers in Python. We have also seen how to evaluate the performance of these models using metrics such as accuracy, precision, recall, and ROC AUC. We have seen that the Random Forest and XGBoost models perform better than the Decision Tree model. Furthermore, we looked at the feature and permutation importance of each model to see which features are most important for determining whether a transaction is fraudulent or not.\n" ] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.8.19" } }, "nbformat": 4, "nbformat_minor": 4 }