Table 2. A detailed comparison of before and after feature selection across different evaluation metrics between the proposed method and other classifiers.

Values separated by “/” indicate “without/with” feature selection.

Model	Sensitivity	Specificity	Precision	F1-Score	Accuracy (%)	Brier score	AUC score
XGBoost	0.28 / 1	0.84 / 0.95	0.80 / 0.97	0.41 / 0.98	53.27 / 98.18	0.479 / 0.023	0.74 / 0.99
Random Forest	0.74 / 0.81	0.61 / 0.75	0.81 / 0.86	0.76 / 0.82	70.72 / 78.54	0.197 / 0.146	0.73 / 0.86
Decision Tree	0.71 / 0.79	0.41 / 0.36	0.66 / 0.68	0.68 / 0.72	60.54 / 62.90	0.392 / 0.372	0.57 / 0.57
Support Vector Machine	0.81 / 1	0.73 / 0.83	0.84 / 0.92	0.81 / 0.96	78.36 / 94.36	0.155 / 0.053	0.84 / 0.97
Naïve Bayes	0.77 / 1	0.31 / 0.83	0.64 / 0.92	0.68 / 0.96	60.90 / 94.36	0.392 / 0.056	0.52 / 0.98
Logistic Regression	0.88 / 1	0.68 / 0.83	0.82 / 0.92	0.84 / 0.96	80.36 / 94.36	0.143 / 0.047	0.82 / 0.98

Notes.

Values separated by ”/” indicate “without/with” feature selection.