Table 4 The evaluation metrics for each model using feature subsets on external test set
Model | Balanced accuracy | Recall | Precision | F1 score |
|---|---|---|---|---|
Decision Tree | 83.38%(71.68–92.42%) | 85.71%(62.50–100.00%) | 40.00%(22.73–58.34%) | 54.55%(35.29-70.00%) |
SVM | 79.92%(67.63–92.28%) | 71.43%(44.44–93.77%) | 47.62%(25.00-69.23%) | 57.14%(33.33–75.57%) |
Random Forest | 84.43%(72.83–92.86%) | 85.71%(64.70–100.00%) | 42.86%(25.92-60.00%) | 57.14%(37.50-74.42%) |
XGBoost | 81.50%(68.90–93.50%) | 71.43%(44.43–92.86%) | 55.56%(30.75-80.00%) | 62.50%(37.02–78.79%) |
AdaBoost | 78.05%(63.99–91.39%) | 57.14%(31.25–81.86%) | 88.89%(64.27–100.00%) | 69.57%(42.11–88.01%) |