Table 6 Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy between (A) original model, (B) original model with rbs kernel adjustment, (C) model B with three additional features, and (D) features in model C with a random forests approach.
PPIs | AUC | AUC 95% CI | Accuracy (%) | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) |
|---|---|---|---|---|---|---|---|
A | 0.9263 | 0.9213–0.9312 | 90.63 | 93.03 | 88.50 | 87.81 | 92.86 |
B | 0.9610 | 0.9577–0.9642 | 92.17 | 92.89 | 91.44 | 90.88 | 92.96 |
C | 0.9629 | 0.9597–0.9660 | 92.24 | 92.17 | 92.30 | 91.70 | 92.32 |
D | 0.9680 | 0.9651–0.9710 | 93.76 | 94.11 | 93.46 | 93.09 | 93.99 |