Fig. 5.

Precision for different subsets of features for the tiles b234, b261, b278, and b360 for a fixed recall (∼0.9). From left to right: the first panel includes the total set of features; in the second, only the magnitude- and period-based features is used; in the third, the magnitude- and extinction-based ones; and, finally, the extinction- and period-based features are used. In all cases, the rows indicate which tile was used for train the RF classifier and the columns indicate the one used for the test.
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.