Fig. 5.

Download original image
Precision-recall curves evaluated on the test sets for the fiducial balanced model (top panel), optimistic balanced (middle panel), and pessimistic balanced (bottom panel). The colors indicate the results we obtained for different folds. F1 is reported as a scoring parameter for each fold and has been evaluated by applying a threshold equal to 0.5 on p(j ∈ k|x, dj) (see Sect. 2.5 for details). The dashed horizontal red line represents the precision-recall curve of a classifier making random guesses.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.