Table 2.
The ROC-AUC and brier loss score for the development, validation and MIMI-IV datasets.
| Classifier | ROC-AUC | Brier Loss | ||||
|---|---|---|---|---|---|---|
| Development | Ex-Validation | MIMIC-IV | Development | Ex-Validation | MIMIC-IV | |
| Logistic Regression | 0.8355 | 0.6775 | 0.7450 | 0.1411 | 0.2016 | 0.1873 |
| Support Vector Machine | 0.8269 | 0.6415 | 0.7214 | 0.1454 | 0.2131 | 0.1738 |
| KNeighborsClassifier | 0.8222 | 0.6581 | 0.6942 | 0.1590 | 0.1872 | 0.1900 |
| DecisionTreeClassifier | 0.5993 | 0.5947 | 0.5420 | 0.3079 | 0.3434 | 0.3651 |
| RandomForestClassifier | 0.8295 | 0.6691 | 0.7317 | 0.1587 | 0.1829 | 0.1750 |
| GaussianNB | 0.7916 | 0.5817 | 0.6814 | 0.2476 | 0.2782 | 0.2603 |
| GradientBoostingClassifier | 0.8224 | 0.6491 | 0.6860 | 0.1565 | 0.2225 | 0.1846 |
| XGBClassifier | 0.8384 | 0.6663 | 0.7167 | 0.1446 | 0.2006 | 0.1718 |
| LGBMClassifier | 0.8333 | 0.6606 | 0.6992 | 0.1746 | 0.2432 | 0.2136 |
| CatBoostClassifier | 0.8455 | 0.6706 | 0.7429 | 0.1451 | 0.1920 | 0.1657 |
| AdaBoostClassifier | 0.8302 | 0.6722 | 0.6518 | 0.2355 | 0.2385 | 0.2333 |
| ExtraTreeClassifier | 0.8327 | 0.6799 | 0.7354 | 0.1546 | 0.1807 | 0.1771 |