ADS Capstone Chronicles Revised

22

Table 4 Advanced Model Accuracy Scores - Validation

Table 5 Advanced Model Accuracy Scores - Test

The results presented in these tables highlight each model’s performance. The Bagging classifier emerged as the most effective model, achieving the highest accuracy of 0.90 in both the validation and test datasets. This indicates its robust ability to detect patterns of fraud and abuse effectively. The Bagging classifier distinguished itself as the top-performing model, delivering

exceptional results across multiple metrics. Its high validation and test accuracies, coupled with an outstanding ROC-AUC score of 0.97 (see Figure 8), underscore its effectiveness in identifying patterns indicative of fraud and abuse. The Bagging Classifier’s ability to combine predictions from multiple decision trees allows it to

manage class imbalance and reduce overfitting, making it particularly well-suited for this project.

146