Prediction of Breast Cancer from Imbalance Respect Using Cluster-Based Undersampling Method

Table 3

Performance comparison based on Wbcd and Bcwo datasets.

Dataset	Method	Accuracy	Sensitivity	Specificity	G-mean	AUC	MCC
Wbcd	K + boosted C5.0	0.982	0.9375	1	0.9682	0.969	0.956
	SMOTEBoost	0.964	0.946	0.978	0.9619	0.963	0.924
	RUSBoost	0.944	0.93	0.954	0.942	0.942	0.886
	SMOTE-Boosted C5.O	0.925	0.939	0.911	0.9248	0.925	0.847
Bcwo	K + boosted C5.0	0.975	0.991	0.969	0.9799	0.980	0.936
	SMOTEBoost	0.92	0.98	0.89	0.934	0.933	0.839
	RUSBoost	0.936	0.926	0.944	0.9350	0.934	0.8539
	SMOTE-Boosted C5.0	0.937	0.934	0.941	0.9375	0.937	0.8756