Table 8.

The classification results (specific features and error rate: SVM, NB, and DA) of the optimal feature combination using leave-one-out validation.

	The classification error rate of the optimal feature combination using leave-one-out validation
	Feature	AVG error rate (SVM,NB,DA)
NSGA	δ:EB-MD-MZ-HR,θ:AT,α:AT-MZ-HR,β:DG-SPL-GE,γ:DG-GIC	0.0624 (0.0122,0.1687,0.0061)
MOPSO	δ:MZ,θ: MZ,α: DG-AT-MD-MZ-HR,β:DG-GE-AT-MZ,γ: CC-SPL-EB-LE-MZ	0.0619 (0.0122,0.1478,0.0258)
MOPSO-M	δ:MZ,θ:MD,α:DG-AT-MZ-HR,β:DG-SPL-GE-AT-MZ,γ: DG-SPL-GE-MZ	0.0593 (0.0092,0.1533,0.0153)
MOPSO-GDM	δ:MZ,θ:AT,α:DG-AT-MZ-HR,γ:DG-MZ	0.0521 (0.0122,0.0981,0.0460)