Table 2 The performance of machine learning models on training sets.
From: Predicting mortality and risk factors of sepsis related ARDS using machine learning models
Cohort and models | AUROC | Cutoff | Sensitivity | Specificity | PPV | NPV |
|---|---|---|---|---|---|---|
XGBoost | 0.951(0.942–0.961) | 0.33 | 0.848(0.82–0.876) | 0.912(0.898–0.925) | 0.784(0.754–0.814) | 0.941(0.929–0.952) |
LightGBM | 1.0(1.0–1.0) | 0.505 | 1.0(1.0–1.0) | 1.0(1.0–1.0) | 1.0(1.0–1.0) | 1.0(1.0–1.0) |
RF | 1.0(1.0–1.0) | 0.462 | 1.0(1.0–1.0) | 1.0(1.0–1.0) | 1.0(1.0–1.0) | 1.0(1.0–1.0) |
NB | 0.793(0.772–0.814) | 0.911 | 0.75(0.716–0.783) | 0.72(0.699–0.741) | 0.504(0.472–0.535) | 0.884(0.867-0.9) |
CART | 0.831(0.811–0.852) | 0.269 | 0.667(0.63–0.703) | 0.897(0.883–0.911) | 0.71(0.674–0.746) | 0.877(0.861–0.892) |
LR | 0.835(0.817–0.854) | 0.335 | 0.684(0.648–0.719) | 0.837(0.82–0.855) | 0.614(0.578–0.649) | 0.875(0.859–0.891) |
