Table 2.

Classification performance of multiple machine learning models.

Model	AUC	Accuracy	Sensitivity/Recall	Specificity	F1-score	PPV/precision	NPV
LR	0.776 | 0.781	0.797 | 0.799	0.633 | 0.657	0.834 | 0.831	0.537 | 0.549	0.465 | 0.471	0.909 | 0.914
NB	0.755 | 0.758	0.821 | 0.832	0.577 | 0.606	0.877 | 0.883	0.545 | 0.572	0.517 | 0.542	0.901 | 0.907
DT	0.984 | 0.829	0.935 | 0.868	0.755 | 0.584	0.976 | 0.933	0.813 | 0.623	0.880 | 0.667	0.946 | 0.907
GB	0.918 | 0.876	0.902 | 0.882	0.768 | 0.730	0.932 | 0.917	0.744 | 0.697	0.721 | 0.667	0.946 | 0.937
RF	0.983 | 0.884	0.935 | 0.871	0.828 | 0.679	0.960 | 0.915	0.826 | 0.662	0.825 | 0.646	0.961 | 0.926
MLP	0.930 | 0.876	0.890 | 0.872	0.768 | 0.745	0.918 | 0.902	0.723 | 0.685	0.682 | 0.634	0.945 | 0.939
XGB	0.982 | 0.871	0.935 | 0.870	0.831 | 0.664	0.958 | 0.917	0.826 | 0.655	0.820 | 0.645	0.961 | 0.923
LGBM	0.979 | 0.886	0.931 | 0.887	0.818 | 0.723	0.957 | 0.925	0.816 | 0.705	0.813 | 0.688	0.958 | 0.936
KNN	0.958 | 0.855	0.900 | 0.840	0.837 | 0.730	0.915 | 0.865	0.806 | 0.615	0.692 | 0.552	0.961 | 0.933

This table demonstrates the predictive performance of various machine learning models (Logistic Regression, Naive Bayes, Decision Tree, Gradient Boosting, Random Forest, Multi-Layer Perceptron, XGBoost, LightGBM, and K-Nearest Neighbors) on both training and validation/test sets. Performance metrics include Area Under the Receiver Operating Characteristic Curve (AUC), accuracy, sensitivity/recall, specificity, F1-score, positive predictive value (PPV/precision), negative predictive value (NPV), and optimal classification cutoff values, enabling comprehensive comparison of model discrimination ability.