Table 3.

The classification performance (in percentage) of four machine learning methods on WM.

			WM
	RetainedFeas	OptimalFeas	Sen	Spec	Acc	AUC
SPEC	9,922	6,048	72.18 ± 3.77	69.35 ± 3.54	70.39 ± 2.33	70.77 ± 3.65
ReliefF	9,922	5,457	74.87 ± 3.45	68.77 ± 2.95	71.18 ± 2.09	71.82 ± 3.28
RFE	7,937	3,360	69.85 ± 4.24	66.31 ± 4.08	68.16 ± 3.49	68.08 ± 4.16
STABLASSO	5,073	3,233	69.30 ± 4.28	71.93 ± 2.25	70.11 ± 2.90	70.62 ± 3.87

The best performance for each indicator is shown in bold.