J/PASJ/72/89ivo://CDS.VizieR/J/PASJ/72/89doi:10.26093/cds/vizier/bj-28-2sCDSTakahashi I.Suzuki N.Yasuda N.Kimura A.Ueda N.Tanaka M.,Tominaga N.Yoshida N.2026-03-17T20:05:50Z2022-04-20T08:11:18ZCDS support team

CDS, Observatoire de Strasbourg, 11 rue de l'Universite, F-67000 Strasbourg, France

cds-question@unistra.frsupernovaeThe advancement of technology has resulted in a rapid increase in supernova (SN) discoveries. The Subaru/Hyper Suprime-Cam (HSC) transient survey, conducted from fall 2016 through spring 2017, yielded 1824 SN candidates. This gave rise to the need for fast type classification for spectroscopic follow-up and prompted us to develop a machine learning algorithm using a deep neural network with highway layers. This algorithm is trained by actual observed cadence and filter combinations such that we can directly input the observed data array without any interpretation. We tested our model with a dataset from the LSST classification challenge (Deep Drilling Field). Our classifier scores an area under the curve (AUC) of 0.996 for binary classification (SN Ia or non-SN Ia) and 95.3% accuracy for three-class classification (SN Ia, SN Ibc, or SN II). Application of our binary classification to HSC transient data yields an AUC score of 0.925. With two weeks of HSC data since the first detection, this classifier achieves 78.1% accuracy for binary classification, and the accuracy increases to 84.2% with the full dataset. This paper discusses the potential use of machine learning for SN type classification purposes.2020PASJ...72...89Thttps://cdsarc.cds.unistra.fr/viz-bin/cat/J/PASJ/72/89CatalogResearchIsServedByTAP VizieR generic servicehttps://cds.unistra.fr/vizier-org/licences_vizier.htmlhttps://vizier.cds.unistra.fr/viz-bin/VizieR-2?-source=J/PASJ/72/89https://vizier.iucaa.in/viz-bin/VizieR-2?-source=J/PASJ/72/89http://vizieridia.saao.ac.za/viz-bin/VizieR-2?-source=J/PASJ/72/89https://vizier.cds.unistra.fr/viz-bin/votable?-source=J/PASJ/72/89https://vizier.iucaa.in/viz-bin/votable?-source=J/PASJ/72/89http://vizieridia.saao.ac.za/viz-bin/votable?-source=J/PASJ/72/89GETtext/xml+votablehttps://tapvizier.cds.unistra.fr/TAPVizieR/tapdefaultJ/PASJ/72/89/table9ML classification results of 1824 HSC SNeIndexIndexmeta.idintrecnoRecord number assigned by the VizieR team. Should Not be used for identification.meta.recordintNameSN name (HSCNNaaaa)meta.id;meta.maincharCase[0/4]? Case (see table6.dat)meta.codeintnullablez? Redshiftsrc.redshiftfloatnullablee_z? Redshift error (lower value)stat.error;src.redshiftfloatnullableE_z? Redshift error (upper value)stat.error;stat.maxfloatnullabler_z[0/4] Redshift source (1)meta.ref;pos.frameintdofDegree of freedom in SALT2 fittingstat.fit.dofintTypeSN type labeled by SALT2 fitting (UC unclassified)meta.notecharFcover[False True -] Flag indicating whether the photometric data covers the period of 10 days before and 20 days after the peakmeta.codecharM2Ia? Ia probability in binary classification with pseudo-absolute magnitude (M) and normalized flux (f)stat.fit.goodnessfloatnullableM3Ia? Ia probability in 3-class classification with pseudo-absolute magnitude (M) and normalized flux (f)stat.fit.goodnessfloatnullableM3Ibc? Ibc probability in 3-class classification with pseudo-absolute magnitude (M) and normalized flux (f)stat.fit.goodnessfloatnullableM3II? II probability in 3-class classification with pseudo-absolute magnitude (M) and normalized flux (f)stat.fit.goodnessfloatnullableM3predSN type in 3-class classification with M and fstat.fit.goodnesscharm2Ia? Ia probability in binary classification with magnitude (m) and normalized flux (f)stat.fit.goodnessfloatnullablem3Ia? Ia probability in 3-class classification with magnitude (m) and normalized flux (f)stat.fit.goodnessfloatnullablem3Ibc? Ibc probability in 3-class classification with magnitude (m) and normalized flux (f)stat.fit.goodnessfloatnullablem3II? II probability in 3-class classification with magnitude (m) and normalized flux (f)stat.fit.goodnessfloatnullablem3predSN type in 3-class classification with magnitude (m) and normalized flux (f)stat.fit.goodnesscharJ/PASJ/72/89/table6Optimized hyperparameters for classificationrecnoRecord number assigned by the VizieR team. Should Not be used for identification.meta.recordintCase[0/4] Casemeta.codeint2TTwo-class hyperparameter T (number of repetitions)stat.fit.paramint2DTwo-class hyperparameter D (number of optimized neurons)stat.fit.paramint2DRTwo-class hyperparameter drop rate (rate of Dropout)stat.fit.paramdouble2bnTwo-class hyperparameter bn (use Batch Normalization if bn=1 else not use Batch Normalization)stat.fit.paramint3TThree-class hyperparameter T (number of repetitions)stat.fit.paramint3DThree-class hyperparameter D (number of optimized neurons)stat.fit.paramint3DRThree-class hyperparameter drop rate (rate of Dropout)stat.fit.paramdouble3bnThree-class hyperparameter bn (use Batch Normalization if bn=1 else not use Batch Normalization)stat.fit.paramint3TypeThree-class hyperparameter type (type of Activation function)stat.fit.paramcharMInput[V-] Pseudo-absolute magnitude (M) as input (yes = V)stat.fit.paramcharmInput[V-] Magnitude (m) as input (yes = V)stat.fit.paramcharfInput[V-] Normalized flux (f) as input (yes = V)stat.fit.paramchar2TypeTwo-class hyperparameter type (type of Activation function)stat.fit.paramchar