Fig. 2: Optimal decision threshold.

a The distribution of the \({\bar{s}}_{({NISSBs},{ISSBs})}\) (blue bars) and \({\bar{s}}_{({PISSBs},{ISSBs})}\) (red bars); b The ROC-AUC curve for the training set.
a The distribution of the \({\bar{s}}_{({NISSBs},{ISSBs})}\) (blue bars) and \({\bar{s}}_{({PISSBs},{ISSBs})}\) (red bars); b The ROC-AUC curve for the training set.