Table 1. Performances of SVM-based models trained with different feature selection methods and features.

Feature selection method	Feature	Sn (%)	Sp (%)	Acc (%)	MCC
NA	DPC	81.34	83.60	81.61	0.64
F-score	ODPC	83.71	86.77	84.84	0.70
Random forest	ODPC	83.67	86.47	84.19	0.70