AGNES-SMOTE: An Oversampling Algorithm Based on Hierarchical Clustering and Improved SMOTE

Table 6

G-mean values on each dataset with different algorithms.

Dataset	SMOTE	K-means-SMOTE	Cluster-SMOTE	AGNES-SMOTE
Ecoli	0.8653	0.8685	0.8518	0.8701
Libra	0.7993	0.6954	0.8478	0.8055
Yeast1	0.7065	0.6596	0.7058	0.7121
Haberman	0.6259	0.5669	0.5909	0.6278
Satimage	0.8519	0.7412	0.8288	0.8396
Optical_digits	0.9689	0.9752	0.9655	0.9753
Abalone	0.7827	0.0875	0.7929	0.7955
Liver	0.6551	0.6407	0.5461	0.6593
LEV	0.7469	0.7920	0.7736	0.8046