TABLE 5. Performance Given Different Feature Sets (AECOPD).
| Feature | Model | Accuracy | Sensitivity | Precision | F1 |
|---|---|---|---|---|---|
| All features | Random forest | 0.921 | 0.904 | 0.943 | 0.923 |
| Environmental & lifestyle features | Decision tree | 0.835 | 0.794 | 0.878 | 0.834 |
| Clinical questionnaire features | LDA | 0.695 | 0.714 | 0.769 | 0.740 |