Imbalanced biomedical data classification using self-adaptive multilayer ELM combined with dynamic GAN

Table 4 Performance comparison results of testing on different datasets

Biomedical datasets	Methods	G-mean	AUC	F-measure
Breast cancer	W-ELM	0.5679	0.8093	0.7658
	SMOTE-ELM	0.7816	0.7981	0.7584
	H-ELM	0.5835	0.6584	0.5967
	PGM-ELM	0.9212	0.9013	0.9354
Liver patient	W-ELM	0.7827	0.7439	0.9127
	SMOTE-ELM	0.6379	0.5198	0.9218
	H-ELM	0.5980	0.6919	0.7226
	PGM-ELM	0.8016	0.8581	0.9304
Diabetic retinopathy	W-ELM	0.6555	0.7849	0.9207
	SMOTE-ELM	0.7554	0.7220	0.7404
	H-ELM	0.6112	0.7493	0.6346
	PGM-ELM	0.8778	0.8619	0.9715
Pima diabetes	W-ELM	0.9360	0.9151	0.9724
	SMOTE-ELM	0.6277	0.8792	0.5655
	H-ELM	0.5041	0.8580	0.5000
	PGM-ELM	0.9657	0.9324	0.9922

ISSN: 1475-925X