Alert Classification for the ALeRCE Broker System: The Anomaly Detector

Simple item page
dc.catalogadorgrr
dc.contributor.authorPérez-Carrasco, Manuel
dc.contributor.authorCabrera-Vives, Guillermo
dc.contributor.authorHernández-García, Lorena
dc.contributor.authorForster, F.
dc.contributor.authorSanchez-Saez, Paula
dc.contributor.authorMuñoz Arancibia, Alejandra M.
dc.contributor.authorArredondo, Javier
dc.contributor.authorAstorga, Nicolas
dc.contributor.authorBauer, Franz Erik
dc.contributor.authorBayo, Amelia
dc.contributor.authorCatelan, Marcio
dc.contributor.authorDastidar, Raya
dc.contributor.authorEstevez, P. A.
dc.contributor.authorLira, Paulina
dc.contributor.authorPignata, Giuliano
dc.date.accessioned2023-09-28T14:44:17Z
dc.date.available2023-09-28T14:44:17Z
dc.date.issued2023
dc.description.abstractAstronomical broker systems, such as Automatic Learning for the Rapid Classification of Events (ALeRCE), are currently analyzing hundreds of thousands of alerts per night, opening up an opportunity to automatically detect anomalous unknown sources. In this work, we present the ALeRCE anomaly detector, composed of three outlier detection algorithms that aim to find transient, periodic, and stochastic anomalous sources within the Zwicky Transient Facility data stream. Our experimental framework consists of cross-validating six anomaly detection algorithms for each of these three classes using the ALeRCE light-curve features. Following the ALeRCE taxonomy, we consider four transient subclasses, five stochastic subclasses, and six periodic subclasses. We evaluate each algorithm by considering each subclass as the anomaly class. For transient and periodic sources the best performance is obtained by a modified version of the deep support vector data description neural network, while for stochastic sources the best results are obtained by calculating the reconstruction error of an autoencoder neural network. Including a visual inspection step for the 10 most promising candidates for each of the 15 ALeRCE subclasses, we detect 31 bogus candidates (i.e., those with photometry or processing issues) and seven potential astrophysical outliers that require follow-up observations for further analysis.
dc.fechaingreso.objetodigital2023-09-27
dc.fuente.origenWOS
dc.identifier.doi10.3847/1538-3881/ace0c1
dc.identifier.eissn1538-3881
dc.identifier.issn0004-6256
dc.identifier.scopusidSCOPUS_ID:85171187981
dc.identifier.urihttps://doi.org/10.3847/1538-3881/ace0c1
dc.identifier.urihttps://repositorio.uc.cl/handle/11534/74732
dc.identifier.wosidWOS:001061049100001
dc.information.autorucInstituto de Astrofísica; Bauer Franz Erik; S/I; 1007961
dc.information.autorucInstituto de Astrofísica; Catelan, Marcio; 0000-0001-6003-8877; 1001556
dc.issue.numero4
dc.language.isoen
dc.nota.accesoContenido completo
dc.pagina.final13
dc.pagina.inicio1
dc.publisherIOP Publishing Ltd
dc.revistaAstronomical Journal
dc.rightsacceso abierto
dc.rights.licenseAtribución 4.0 Internacional (CC BY 4.0)
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/deed.es
dc.subjectAstronomy data analysis
dc.subjectSurveys
dc.subjectInterdisciplinary astronomy
dc.subject.ddc520
dc.subject.deweyAstronomíaes_ES
dc.subject.ods07 Affordable and Clean Energy
dc.subject.odspa07 Energía asequible y no contaminante
dc.titleAlert Classification for the ALeRCE Broker System: The Anomaly Detector
dc.typeartículo
dc.volumen166
sipa.codpersvinculados1007961
sipa.codpersvinculados1001556
sipa.indexWOS
sipa.trazabilidadWOS;2023-09-23
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Perez-Carrasco_2023_AJ_166_151 (1).pdf
Size:
49.7 MB
Format:
Adobe Portable Document Format
Description:
Download
Collections
Artículos de revistas