Table 2.

Evaluation metrics of the models constructed by each method.

	Method	AUC	ACC	PPV	NPV	SEN	SPE	F1 score	MCC	KAPPA	Brier score
Tra	GBM	0.796	0.717	0.546	0.849	0.736	0.708	0.627	0.419	0.407	0.165
	LR	0.76	0.685	0.509	0.837	0.728	0.664	0.599	0.368	0.353	0.178
	Nnet	0.778	0.706	0.534	0.836	0.71	0.705	0.61	0.392	0.382	0.172
	RF	0.96	0.882	0.76	0.962	0.928	0.86	0.835	0.754	0.745	0.084
	SVM	0.8	0.747	0.577	0.887	0.808	0.718	0.674	0.494	0.476	0.166
	XGBoost	0.995	0.968	0.937	0.984	0.967	0.969	0.952	0.929	0.928	0.042
Val	GBM	0.786	0.716	0.539	0.842	0.709	0.719	0.612	0.404	0.395	0.168
	LR	0.767	0.689	0.506	0.847	0.74	0.665	0.601	0.378	0.36	0.175
	Nnet	0.79	0.694	0.511	0.865	0.78	0.654	0.618	0.404	0.38	0.166
	RF	0.979	0.919	0.817	0.98	0.96	0.9	0.882	0.828	0.821	0.068
	SVM	0.837	0.772	0.597	0.921	0.866	0.728	0.706	0.554	0.53	0.154
	XGBoost	1	0.999	0.998	0.999	0.998	0.999	0.998	0.997	0.997	0.013

Tra, training set; Val, validation set; AUC, area under the curve; ACC, accuracy; PPV, positive predictive value; NPV, negative predictive value; SEN: sensitivity; SPE: specificity, MCC, Matthews correlation coefficient.