A feature selection method based on multiple kernel learning with expression profiles of different types

Table 3 The results of mean effectiveness on mRNA microarray (top 10)

Methods	SVM-RFE	SVM-RCE	mRMR	IMRelief	SlimPLS	OSFS	FGM	SMKL-FS
Liver	0.913	0.860	0.965	0.825	0.831	0.750	0.867	0.963
Pancreatic	0.689	0.777	0.818	0.784	0.673	0.707	0.729	0.804
Lung	0.731	0.786	0.942	0.814	0.708	0.704	0.860	0.964
Gastric	0.614	0.724	0.688	0.566	0.636	0.533	0.640	0.760
Colon	0.736	0.888	0.941	0.803	0.794	0.682	0.812	0.951
Breast	0.745	0.776	0.832	0.545	0.693	0.728	0.769	0.854
Thyroid	0.835	0.897	0.838	0.633	0.743	0.517	0.802	0.922
Prostate	0.577	0.762	0.750	0.560	0.682	0.629	0.679	0.717
Mean	0.730	0.809	0.847	0.691	0.720	0.656	0.770	0.867

ISSN: 1756-0381