AUC (between-datasets setting).

AUC obtained with Nearest Centroids when a signature is learnt from one dataset and tested by 10-fold cross-validation on the three remaining datasets. Standard error is shown within parentheses. For each training dataset, we highlighted the best performance. The Type column refers to the use of feature selection run a single time (S) or through ensemble feature selection, either with the mean (E-M), exponential (E-E) or stability selection (E-S) procedure to aggregate lists.