Refining breast cancer biomarker discovery and drug targeting through an advanced data-driven approach

BMC Bioinformatics

Table 3 Comparison of the performance metrics of different feature selection methods

Algorithm	Dataset	NF	F1	PR-AUC	ROC-AUC	MCC	BAc
All Features	Merged Dataset	10629	0.862	0.870	0.858	0.814	0.905
All Features	GSE45827	11731	0.873	.978	0.961	0.895	0.795
BGWO_SA_Ens	Merged Dataset	1404	0.984	0.986	0.977	0.967	0.977
BGWO_SA_Ens	GSE45827	1710	0.965	0.986	0.972	0.953	0.965
BGWO_Ens	Merged Dataset	1941	0.898	0.972	0.965	0.839	0.916
BGWO_Ens	GSE45827	2114	0.865	0.965	0.931	0.829	0.914
GA_Ens	Merged Dataset	2557	0.954	0.986	0.977	0.943	0.965
GA_Ens	GSE45827	2488	0.908	0.986	0.972	0.894	0.947
LASSO	Merged Dataset	10	0.885	0.874	0.852	0.831	0.913
LASSO	GSE45827	8	0.729	0.904	0.872	0.571	0.772
MCFS-IFS	Merged Dataset	1850	0.940	0.979	0.973	0.945	0.898
MCFS-IFS	GSE45827	475	0.951	0.975	0.963	0.930	0.866
mRMR-IFS	Merged Dataset	355	0.945	0.971	0.954	0.916	0.953
mRMR-IFS	GSE45827	1660	0.926	0.940	0.964	0.897	0.934

ISSN: 1471-2105