Performance metrics of both models in the aggregated cross-validation sets and the test set
| Model | Modelling phase | AUROC | AUPRC | Brier score* | F1-score† |
| Gradient boosted trees | Cross-validation mean | 0.77 (SD=0.03) | 0.340 | 0.066 | 0.16 |
| Test | 0.77 (95% CI 0.73 to 0.82) | 0.37 | 0.092 | 0.17 | |
| Logistic regression | Cross-validation mean | 0.75 (SD=0.02) | 0.31 | 0.098 | 0.14 |
| Test | 0.78 (95% CI 0.73 to 0.82) | 0.37 | 0.092 | 0.16 |
*The Brier score is a cost function that measures performance of probabilistic predictions. The score ranges from 0 to 1. The lower the score, the more accurate the prediction.
†F1-scores present a balance between precision and recall. The higher the score, the more accurate the prediction.
AUPRC, area under the precision recall curve; AUROC, area under the curve of the receiver operating characteristics.