TABLE 4.

Comparison of multiple machine learning methods using fusion features.

Cross-validation	Feature	ML method	AUROC	Sn	Sp	Acc	Mcc
10-fold cross-validation	Fusion	SVM	0.67	0.51	0.80	0.66	0.34
	Fusion	LightGBM	0.92	0.85	0.85	0.85	0.70
	Fusion	DT	0.80	0.83	0.77	0.80	0.60
	Fusion	LR	0.82	0.74	0.77	0.76	0.52
	Fusion	RF	0.93	0.86	0.84	0.85	0.70
Independent set validation	Fusion	SVM	0.74	0.61	0.78	0.70	0.40
	Fusion	LightGBM	0.97	0.92	0.91	0.91	0.83
	Fusion	DT	0.94	0.94	0.84	0.89	0.78
	Fusion	LR	0.89	0.80	0.84	0.82	0.64
	Fusion	RF	0.98	0.94	0.94	0.94	0.88

Best performance metrics are shown in bold.