Skip to main content
Fig. 5 | BioData Mining

Fig. 5

From: A self-inspected adaptive SMOTE algorithm (SASMOTE) for highly imbalanced data classification in healthcare

Fig. 5

For case 2: A F1 scores, B precisions and C recalls of the CHD risk prediction models built on SASMOTE, SASMOTE without invisible neighbors (SASMOTE w/o visible), SASMOTE without inspections (SASMOTE w/o inspections), B-SMOTE, SMOTE, and without data resampling under different balanced ratios. The proposed SASMOTE performs better on the average F1 score under most imbalanced ratios. The B-SMOTE performs best on the average precision. The original dataset performs best on the average recall, but worst on the precision and F1 score

Back to article page

BioData Mining

ISSN: 1756-0381