Table 12 Performance metrics for comprehensive ablation study across datasets.

From: Optimizing imbalanced learning with genetic algorithm

Dataset

Metric

No crossover

No mutation

LR init

Random init

Full SVM-GA

Credit card fraud

F1-score

0.70

0.74

0.78

0.73

0.82

Precision

0.67

0.72

0.76

0.71

0.82

ROC AUC

0.84

0.82

0.88

0.85

0.92

PIMA diabetes

F1-score

0.62

0.61

0.68

0.64

0.71

Precision

0.50

0.53

0.56

0.52

0.59

ROC AUC

0.65

0.69

0.73

0.70

0.76

Phoneme

F1-score

0.65

0.68

0.70

0.67

0.72

Precision

0.49

0.51

0.57

0.53

0.60

ROC AUC

0.72

0.71

0.78

0.75

0.82

Back to article page