Transient classifiers for Fink - Benchmarks for LSST

Open Access

Table 1

CATS results for the validation sample across all folds.

	ROC	PR	Precision	Recall
	AUC	AUC	Precision	Recall
SN-like	0.99	0.99	0.97	0.99
	(0.0001)	(0.0002)	(0.0022)	(0.0013)
Fast	0.99	0.80	0.85	0.72
	(0.0013)	(0.047)	(0.092)	(0.017)
Long	0.96	0.68	0.82	0.47
	(0.0012)	(0.0041)	(0.032)	(]0.041)
Periodic	1.00	1.00	1.00	1.00
	(0.00001)	(0.00003)	(0.0002)	(0.0001)
Non-Periodic	1.00	1.00	0.97	0.96
	(0.00003)	(0.0005)	(0.003)	(0.005)

Notes. Reported are the means with standard deviations in parentheses. ROC AUC and PR AUC are the area under the curve for the ROC and precision-recall curves, respectively, while the values for the precision and recall are taken from the confusion matrix averaged over all folds.

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.