Table 2.
A comparison of AUC, Sensitivity+, and Specificity+ for Gaussian Naïve Bayes classifier using alphacarbon, averagecx, rgroup, and roughness features respectively, Naïve Bayes using sequence features, and Struct-NB-Ensemble that combines sequence and structural features. The values are reported for the same threshold θ = 0.5 on the ROC curve.
| Name | AUC | Sens+ | Spec+ |
|---|---|---|---|
| NB-AC | 0.658 | 0.368 | 0.36 |
| NB-ACX | 0.672 | 0.397 | 0.366 |
| NB-RG | 0.679 | 0.403 | 0.367 |
| NB-RN | 0.696 | 0.469 | 0.349 |
| NB-Seq | 0.736 | 0.246 | 0.65 |
| Struct-NB-Ensemble | 0.752 | 0.359 | 0.443 |