Table 9.

Task performance and ranking quality for CWR dataset.

Approach	RQS	$\mathit{m}\mathit{a}\mathit{x}\left(\mathit{P}\mathit{M}\right)$	Number of Variables for $\mathit{m}\mathit{a}\mathit{x}\left(\mathit{P}\mathit{M}\right)$	PM for 95% Threshold	Number of Features for 95% Threshold
FS layer	0.9754	1.0014	44	0.9599	29
Mutual Info.	0.7215	1.0000	100	0.9626	85
ReliefF	0.8630	1.0000	95	0.9587	63
Random Forest	0.8346	1.0000	98	0.9646	80
AFS	0.7127	0.9997	100	0.9629	92
AFS 2	0.7466	0.9997	100	0.9545	87