Skip to main content
. 2023 Jun 21;14:1158555. doi: 10.3389/fneur.2023.1158555

Table 5.

Comparison of the classification performance of various balancing methods (%).

Method Type Macro-AUC ACC Macro-R Macro-P Macro-F1
Original None 92.60 81.16 85.01 74.62 78.75
Random Oversample Oversampling 91.84 80.37 80.37 81.14 80.49
Random Undersample Oversampling 82.78 73.81 71.82 74.52 72.47
SMOTE Undersampling 90.98 85.93 86.10 85.85 85.85
ADASYN Undersampling 95.74 84.84 85.05 85.35 84.28
Borderline-SMOTE Undersampling 96.10 87.04 87.12 87.68 86.75
SMOTEENN Mixed-sampling 97.89 95.74 94.75 94.03 94.35