Table 5. Statistics for machine learning models with the broader set of inputs and for the BCRAT.
| AUC | Sensitivity | Specificity | Precision | |
|---|---|---|---|---|
| LR | 0.613 (0.579-0.647 95% CI) | 0.476 (0.416-0.536 95% CI) | 0.691 (0.683-0.699 95% CI) | 0.0323 (0.0285-0.0366 95% CI) |
| NB | 0.589 (0.555-0.623 95% CI) | 0.639 (0.582-0.697 95% CI) | 0.523 (0.514-0.531 95% CI) | 0.0282 (0.0258-0.0308 95% CI) |
| DT | 0.508 (0.496-0.521 95% CI) | 0.0446 (0.0199-0.0693 95% CI) | 0.972 (0.969-0.974 95% CI) | 0.0328 (0.0190-0.0562 95% CI) |
| LDA | 0.613 (0.579-0.646 95% CI) | 0.688 (0.632-0.743 95% CI) | 0.467 (0.459-0.476 95% CI) | 0.0272 (0.0251-0.0295 95% CI) |
| SVM | 0.518 (0.484-0.551 95% CI) | 0.517 (0.457-0.576 95% CI) | 0.478 (0.469-0.486 95% CI) | 0.0210 (0.0187-0.0235 95% CI) |
| NN | 0.608 (0.574-0.643 95% CI) | 0.599 (0.540-0.657 95% CI) | 0.562 (0.553-0.570 95% CI) | 0.0287 (0.0261-0.0317 95% CI) |
| BCRAT | 0.563 (0.528-0.597 95% CI) | 0.647 (0.590-0.704 95% CI)* | 0.461 (0.452-0.470 95% CI)* | 0.0254 (0.0232-0.0277 95% CI)* |
* Calculated using sensitivity / specificity values based on the threshold that maximized the sum of testing data set sensitivities and specificities rather than the sum of training data set sensitivities and specificities.