Table 3.
Model performance of four different algorithms.
| Methods | AUCa | Accuracy | Recall | F1-score | Specificity |
| Logistic regression | 0.8544 | 0.7726 | 0.7141 | 0.7354 | 0.8191 |
| SVMb | 0.8898 | 0.8112 | 0.7844 | 0.7861 | 0.8325 |
| Random forest | 0.8956 | 0.8343 | 0.8157 | 0.8133 | 0.8490 |
| XGBoostc | 0.9220 | 0.8478 | 0.8512 | 0.8319 | 0.8451 |
aAUC: area under the receiver operating curve.
bSVM: support vector machine.
cXGBoost: extreme gradient boosting.