Table 4.
Predictive performance of each model for predicting HBV infection risk.
| Algorithms | AUC | Standard Error | 95% CI | AUC Compared with LR |
|---|---|---|---|---|
| LR | 0.742 | 0.006 | (0.729, 0.754) | - |
| DT | 0.619 | 0.008 | (0.603, 0.634) | −0.123 |
| RF | 0.752 | 0.006 | (0.740, 0.764) | +0.010 |
| XGBoost | 0.779 | 0.006 | (0.768, 0.791) | +0.037 |
| Borderline-SMOTE DT | 0.715 | 0.007 | (0.702, 0.729) | −0.027 |
| Borderline-SMOTE RF | 0.759 | 0.006 | (0.747, 0.771) | +0.017 |
| Borderline-SMOTE XGBoost | 0.782 | 0.006 | (0.771, 0.793) | +0.040 |
LR: logistic regression; SMOTE: synthetic minority oversampling technique; AUC: the area under the receiver operating characteristic curve; CI: confidence interval.