Table 1. Incremental contribution of the features used in our model for the predictive quality.
Scheme | Sen. | Spe. | MCC | AUC (S.D.) |
---|---|---|---|---|
DEO | 0.63 | 0.68 | 0.31 | 0.64 (0.01) |
DEO+ESS | 0.63 | 0.70 | 0.33 | 0.65 (0.01) |
DEO+REC | 0.66 | 0.61 | 0.27 | 0.66 (0.01) |
DEO+ESS+REC | 0.72 | 0.66 | 0.38 | 0.74 (0.01) |
DEO+ESS+REC+PATH | 0.71 | 0.70 | 0.42 | 0.79 (0.01) |
Sen and Spe refer respectively to sensitivity and specificity. MCC and AUC are the Matthew correlation coefficient and Area Under the Curve results.