Multi-Nyström Method Based on Multiple Kernel Learning for Large Scale Imbalanced Classification

Table 2

F1 score, G-mean, and AUC results of different algorithms on four datasets

Datasets	Measures	SVM	Nyström	Multi-Nyström	MKSVM
Poker-8-9_vs_5	F1	0.0571	0.0327	0.0585	0.1906
	G-mean	0.0399	0.3357	0.5140	0.1589
	AUC	0.8107	0.6106	0.7953	0.7942
Abalone19	F1	0.0611	0.0200	0.0334	0.0569
	G-mean	0.0661	0.2914	0.3500	0.0661
	AUC	0.7487	0.5203	0.6094	0.7263
Page-blocks0	F1	0.8061	0.7954	0.8171	0.8342
	G-mean	0.7018	0.7292	0.7115	0.7381
	AUC	0.9857	0.9753	0.9585	0.9904
USPS	F1	0.8991	0.6688	0.8853	0.9102
	G-mean	0.8788	0.8593	0.8408	0.8807
	AUC	0.9939	0.9608	0.9874	0.9963