TABLE 9. Performance Comparison of Various Machine Learning Models for Test Set.
| SVM (Linear) | SVM (RBF) | RF | GBDT | |
|---|---|---|---|---|
| Accuracy | 70.4% | 71.0% | 69.5% | 68.9% |
| Specificity | 70.2% | 70.8% | 70.5% | 70.8% |
| Sensitivity | 72.2% | 72.2% | 61.1% | 52.8% |
| Precision | 22.2% | 22.6% | 19.6% | 17.6% |
| F1-score | 34.0% | 34.4% | 29.7% | 26.4% |
| ROC AUC | 0.777 | 0.780 | 0.710 | 0.683 |
| PR AUC | 0.295 | 0.297 | 0.300 | 0.330 |
| True Negative | 214 | 216 | 215 | 216 |
| False Negative | 10 | 10 | 14 | 17 |
| False Positive | 91 | 89 | 90 | 89 |
| True Positive | 26 | 26 | 22 | 19 |