Table 6 The performance metrics of the proposed method.
Method | Metric | Mean | ± SD | 95% confidence interval | p-value | Statistically significant |
|---|---|---|---|---|---|---|
Deep-CTGAN + ResNet + TabNet | Accuracy | 83.4% | ± 1.1 | [81.83%, 84.97%] | 0.005 | Yes |
F1-score | 0.81 | ± 1.1 | [0.796, 0.824] | 0.005 | Yes | |
AUC | 0.85 | ± 1.1 | [0.836, 0.864] | 0.005 | Yes | |
SMOTE + TabNet | Accuracy | 79.3% | ± 1.3 | [77.69%, 80.91%] | 0.015 | Yes |
F1-score | 0.76 | ± 1.3 | [0.742, 0.778] | 0.015 | Yes | |
AUC | 0.79 | ± 1.3 | [0.772, 0.808] | 0.015 | Yes | |
ADASYN + TabNet | Accuracy | 78.7% | ± 1.4 | [76.96%, 80.44%] | 0.020 | Yes |
F1-score | 0.75 | ± 1.4 | [0.730, 0.770] | 0.020 | Yes | |
AUC | 0.78 | ± 1.4 | [0.760, 0.800] | 0.020 | Yes | |
Deep-CTGAN + TabNet | Accuracy | 81.6% | ± 1.2 | [80.13%, 83.07%] | 0.009 | Yes |
F1-score | 0.79 | ± 1.2 | [0.774, 0.806] | 0.009 | Yes | |
AUC | 0.82 | ± 1.2 | [0.804, 0.836] | 0.009 | Yes | |
ResNet + TabNet | Accuracy | 80.2% | ± 1.3 | [78.59%, 81.81%] | 0.012 | Yes |
F1-score | 0.77 | ± 1.3 | [0.742, 0.798] | 0.012 | Yes | |
AUC | 0.80 | ± 1.3 | [0.782, 0.818] | 0.012 | Yes |
