JPPRED: Prediction of Types of J-Proteins from Imbalanced Data Using an Ensemble Learning Method

Table 2

Performance of ensemble learning method and individual base classifiers by 10-fold cross validation.

Subfamily	Measure	Ensemble	RF	NB	LR	RBF network
Type I J-protein	Sn	0.905	0.698	0.857	0.556	0
	Sp	0.849	0.935	0.882	0.853	0.934
Type II J-protein	Sn	0.745	0.418	0.691	0.400	0
	Sp	0.857	0.947	0.890	0.858	0.927
Type III J-protein	Sn	0.851	0.979	0.895	0.889	1
	Sp	0.855	0.486	0.768	0.442	0
Type IV J-protein	Sn	1	0	0.700	0.200	0
	Sp	0.849	0.938	0.884	0.848	0.900
AvgSn		0.875	0.524	0.786	0.511	0.250
Acc		0.852	0.922	0.881	0.837	0.885