Table 5 Six algorithms’ model performance in the test set
Models | Sensitivity | Specificity | Precision | Recall | F1 score | AUC |
|---|---|---|---|---|---|---|
XGBoost | 0.678 | 0.682 | 0.360 | 0.678 | 0.470 | 0.720 (0.661–0.779) |
RF | 0.724 | 0.667 | 0.364 | 0.724 | 0.484 | 0.763 (0.710–0.815) |
CatBoost | 0.747 | 0.658 | 0.205 | 0.977 | 0.339 | 0.743 (0.689–0.797) |
GBDT | 0.736 | 0.679 | 0.376 | 0.736 | 0.498 | 0.761 (0.707–0.816) |
SVM | 0.667 | 0.688 | 0.360 | 0.667 | 0.468 | 0.677 (0.622–0.733) |
Logistic regression | 0.598 | 0.748 | 0.382 | 0.600 | 0.467 | 0.697(0.633–0.762) |