Fig. 5From: Data reduction techniques for highly imbalanced medicare Big DataMethodology scenario three experiments, for feature selection, then RUSBack to article page