Table 4 Average model performance on credit card fraud detection dataset.

From: Optimizing imbalanced learning with genetic algorithm

Metric

SMOTE9

ADASYN11

GAN71

VAE63

SGA

EGA

SVMGA

Accuracy

99.9

99.9

99.9

99.9

99.9

99.9

99.9

Precision

70.88

71.98

84.9

84.1

85.12

87.3

83.92

Recall

82.90

82.06

74.7

80.1

81.32

81.32

82.40

F1 score

75.88

76.34

80.5

82.0

83.38

84.04

83.12

ROC AUC

91.30

91.24

87.2

90.0

90.70

90.70

91.48

  1. This table presents the performance metrics (accuracy, precision, recall, F1 score, and ROC AUC) for models trained with different synthetic data generation techniques (SMOTE, ADASYN, GAN, VAE, SGA, EGA, and SVMGA) on the credit card fraud detection dataset.
  2. Significant values are in bold.
Back to article page