Table 2.

Performance comparison of six classification ML models

Characteristics	SVM	LGBM	RF	GBDT	XGBoost	CatBoost
AUC	0.776	0.823	0.794	0.809	0.798	0.776
AUC 95% CI	0.748–0.805	0.798–0.848	0.769–0.818	0.783–0.834	0.773–0.823	0.747–0.804
Accuracy	0.867	0.953	0.933	0.933	0.978	0.984
Precision	0.467	1.000	1.000	1.000	1.000	1.000
Recall	1.000	0.648	0.323	0.719	0.716	0.682
F1 score	0.636	0.786	0.488	0.837	0.835	0.811
Brier score	0.145	0.077	0.077	0.114	0.068	0.061
AP	0.142	0.175	0.145	0.173	0.141	0.147

SVM, support vector machine; LGBM, light gradient boosting machine; RF, random forest; GBDT, gradient boosting decision tree; XGBoost, extreme gradient boosting; CatBoost, category boosting; AUC, the area under the ROC; AP, the area under the P-R curve