Fig. 7

Download original image
Examples of feature importance derived from the RandomForest, CatBoost, and LightGBM classifiers when selecting quiescent galaxies from within the 0 < z < 3 interval, without foreknowledge of galaxy redshifts. For each learner, the feature importance values are normalised such that their sum is 1.0. The y axis labels correspond to feature names used by the pipeline after the pre-processing steps outlined in Sect. 4.2 have been applied and should be self-explanatory. For example, the feature named 'u' is derived from the u-band magnitudes (i.e. after our pre-processing steps), the feature 'VIS-20CM' is derived from the IE − 20 cm colours, etc.
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.