Table 4.
Performance achieved by scikit ML on the 3D descriptors.
| Methods (Parameters) | Main Dataset | Validation Dataset | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Sen | Spc | Acc | MCC | AUC | Sen | Spc | Acc | MCC | AUC | |
| RF (n_estimators = 800) | 60.22 | 66.09 | 63.16 | 0.26 | 0.69 | 58.29 | 65.64 | 61.97 | 0.24 | 0.67 |
| KNN (n_neighbors = 10,algorithm = ‘ball_tree',weights = ‘distance') | 60.43 | 54.94 | 57.68 | 0.15 | 0.61 | 51.28 | 58.97 | 55.13 | 0.1 | 0.59 |
| Ridge (alpha = 0.01) | 58.71 | 60.73 | 59.72 | 0.19 | 0.65 | 49.57 | 63.25 | 56.41 | 0.13 | 0.59 |
| Extratree (n_estimator = 70) | 65.59 | 65.88 | 65.74 | 0.31 | 0.70 | 61.71 | 64.79 | 63.25 | 0.27 | 0.68 |