Optimal Statistical Feature Subset Selection for Bearing Fault Detection and Severity Estimation

Table 10

Classifier performance in terms of F-measure and G-mean values for Exp. 2.

Classifier	F-measure				Geometric mean (G-mean)
	Normal	Inner race	Outer race	Ball bearing	Normal	Inner race	Outer race	Ball bearing
PART	0.990	0.955	0.949	0.932	0.994	0.970	0.969	0.949
PART + multiboost	0.995	0.967	0.958	0.948	0.997	0.977	0.974	0.962
PART + bagging	0.994	0.966	0.957	0.946	0.995	0.976	0.973	0.961
JRIP	0.987	0.931	0.929	0.934	0.992	0.958	0.947	0.951
JRIP + multiboost	0.993	0.965	0.955	0.944	0.995	0.972	0.973	0.961
JRIP + bagging	0.991	0.958	0.957	0.944	0.995	0.971	0.973	0.960