Table 3.

Details of the features used to obtain the optimal feature subset in each phase.

Dataset	Phase 1		Phase 2
	$\mathbf{Number}\mathbf{of}\mathbf{Features}\mathbf{Used}\mathbf{for}\mathbf{Union}\left(\mathit{m}\right)$	$\mathbf{Number}\mathbf{of}\mathbf{Features}\mathbf{Selected}\left(\mathit{k}\right)$	Number of Discarded Correlated Features	Number of Non-Correlated Features	$\mathbf{Number}\mathbf{of}\mathbf{Features}\mathbf{Selected}\left(\mathit{j}\right)$
Arrhythmia	50	60	25	35	20
Leukemia	100	70	46	24	23
DLBCL	150	50	26	24	23
Prostate cancer	200	75	50	25	25