Selection of powerful radio galaxies with machine learning

Home

All issues

Volume 679 (November 2023)

A&A, 679 (2023) A101

Full HTML

Press Release

Open Access

Table 15.

Results of application of several AGN detection criteria to our testing subset and the labelled sources from the S82 field.

HETDEX test set
Method ^(a)	F_β	MCC	Precision	Recall
	(× 100)	(× 100)	(× 100)	(× 100)
S12	86.10	78.78	93.98	80.51
M12	51.80	49.71	98.87	37.18
M16	67.21	61.30	97.48	53.48
B18	82.14	75.76	97.54	72.66
This work	92.71	87.64	94.00	91.67

S82 (labelled)
Method	F_β	MCC	Precision	Recall
	(× 100)	(× 100)	(× 100)	(× 100)

S12	83.59	45.47	93.93	76.62
M12	46.80	28.22	99.59	32.54
M16	64.69	37.76	98.80	50.32
B18	79.71	51.07	98.72	68.77
This work	90.63	58.53	94.15	87.91

Notes.

^(a)

Naming codes for the used methods are described in the main text (cf. Sect. 5.1.1). Last row of each sub-table corresponds to the criterion derived in this work (as described in Sect. 5.3.1). All metrics have been multiplied by 100.

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.