Transient classifiers for Fink - Benchmarks for LSST

Open Access

Table 2

SUPERNNOVA results for a blind test sample.

Class	Accuracy	ROC	Precision	Recall
Class	Accuracy	AUC	Precision	Recall
SN-like	95.58	0.9890	96.19	93.38
Fast	98.15	0.9958	99.98	96.3
Long	84.58	0.9270	83.85	87.76
Periodic	99.93	0.9999	99.98	99.9
Non-Periodic	99.4	0.9999	99.98	99.07
Broad	88.52	–	68.15	80.0

Notes. Results obtained on complete light curves using an independent test set from the training sample. All rows except the last one show the metrics for a binary target versus other types.

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.