A multi-band AGN-SFG classifier for extragalactic radio surveys using machine learning

Open Access

Table B.1

Comparison of the macro F₁ -score performance of the model for different missing-feature strategies.

	Only dropped	Retrained
No missing	0.88	-
LOFAR 144 MHz	0.77	0.84
u	0.86	0.88
J, H, K	0.87	0.88
g, r, i, z, y	0.86	0.88
ch1, ch2	0.83	0.88
ch3, ch4	0.83	0.87
ch1, ch2, ch3, ch4	0.74	0.86
MIPS24	0.85	0.86
PACS100, PACS160	0.87	0.87
SPIRE250, SPIRE350, SPIRE500	0.83	0.85

Notes. The left column shows the model performance that was not retrained after one or multiple features were removed, but the right column was retrained.

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.