An Improved Oversampling Algorithm Based on the Samples’ Selection Strategy for Classifying Imbalanced Data

Table 4

The AUC values of the different methods for the four data sets.

Data sets	Original SVM	SMOTE	Random-SMOTE	AKN-Random-SMOTE
Banana	0.5845	0.6695	0.7378	0.8912
Haberman	0.7560	0.8205	0.8451	0.9727
Pima	0.7491	0.7716	0.8189	0.9718
Led	0.8804	0.9676	0.9910	0.9962