The miniJPAS survey: star-galaxy classification using machine learning

Author(s):Show less -

Baqui, P. O. ^{[1, 2]} ; Marra, V. ^{[3, 4, 1, 2, 5, 6]} ; Casarini, L. ^[7] ; Angulo, R. ^{[8, 9, 10]} ; Diaz-Garcia, L. A. ; Hernandez-Monteagudo, C. ^{[11, 12, 13]} ; Lopes, P. A. A. ^[14] ; Lopez-Sanjuan, C. ^[11] ; Muniesa, D. ^[15] ; Placco, V. M. ^[16] ; Quartin, M. ^{[17, 14]} ; Queiroz, C. ^[18] ; Sobral, D. ^[19] ; Solano, E. ^[20] ; Tempel, E. ^[21] ; Varela, J. ^[11] ; Vilchez, J. M. ^[22] ; Abramo, R. ^[18] ; Alcaniz, J. ^[23] ; Benitez, N. ^[22] ; Bonoli, S. ^{[8, 9, 15]} ; Carneiro, S. ^[24] ; Cenarro, A. J. ^[11] ; Cristobal-Hornillos, D. ^[15] ; de Amorim, A. L. ^[25] ; de Oliveira, C. M. ^[26] ; Dupke, R. ^{[27, 28, 23]} ; Ederoclite, A. ^[26] ; Gonzalez Delgado, R. M. ^[22] ; Marin-Franch, A. ^[11] ; Moles, M. ; Ramio, H. Vazquez ^[11] ; Sodre, L. ^{[26, 15]} ; Taylor, K. ^[29]

Affiliation:Show less -

^[1] Univ Fed Espirito Santo, PPGFis, BR-29075910 Vitoria, ES - Brazil

^[2] Univ Fed Espirito Santo, Nucleo Astrofis & Cosmol Cosmo Ufes, BR-29075910 Vitoria, ES - Brazil

^[3] Univ Fed Espirito Santo, Dept Fis, BR-29075910 Vitoria, ES - Brazil

^[4] INAF Osservatorio Astron Trieste, Via Tiepolo 11, I-34131 Trieste - Italy

^[5] Univ Fed Espirito Santo, PPGCosmo, BR-29075910 Vitoria, ES - Brazil

^[6] IFPU Inst Fundamental Phys Universe, Via Beirut 2, I-34151 Trieste - Italy

^[7] Univ Fed Sergipe, Dept Fis, BR-49100000 Aracaju, SE - Brazil

^[8] Ikerbasque, Basque Fdn Sci, Bilbao 48013 - Spain

^[9] Donostia Int Phys Ctr DIPC, Manuel Lardizabal Ibilbidea 4, San Sebastian - Spain

^[10] AS NTU, Acad Sinica Inst Astron & Astrophys ASIAA, 11F Astron Math Bldg, 1, Sect 4, Roosevelt Rd, Taipei 10617 - Taiwan

^[11] CSIC, Ctr Estudios Fis Cosmos Aragon CEFCA, Unidad Asociada, Plaza San Juan 1, Teruel 44001 - Spain

^[12] Inst Astrofis Canarias, C Via Lactea S-N, Tenerife 38205 - Spain

^[13] Univ La Laguna, Dept Astrofis, Tenerife 38206 - Spain

^[14] Univ Fed Rio de Janeiro, Observ Valongo, BR-20080090 Rio de Janeiro, RJ - Brazil

^[15] Ctr Estudios Fis Cosmos Aragon CEFCA, Plaza San Juan 1, Teruel 44001 - Spain

^[16] NSFs Opt Infrared Astron Res Lab, Tucson, AZ 85719 - USA

^[17] Univ Fed Rio de Janeiro, Inst Fis, BR-21941972 Rio De Janeiro, RJ - Brazil

^[18] Univ Sao Paulo, Inst Fis, BR-05508090 Sao Paulo, SP - Brazil

^[19] Univ Lancaster, Phys Dept, Bailrigg, Leics - England

^[20] CSIC, INTA, Ctr Astrobiol, Dept Astrofis, ESAC Campus, Camino Bajo Castillo S-N, Madrid 28692 - Spain

^[21] Univ Tartu, Tartu Observ, Observ 1, EE-61602 Toravere - Estonia

^[22] CSIC, Inst Astrofis Andalucia, Apdo 3004, Granada 18080 - Spain

^[23] Minist Ciencia Tecnol Inovacao & Comunicacoes, Observ Nacl, BR-20921400 Rio De Janeiro, RJ - Brazil

^[24] Univ Fed Bahia, Inst Fis, BR-40210340 Salvador, BA - Brazil

^[25] Univ Fed Santa Catarina, CFM, Dept Fis, BR-88040900 Florianopolis, SC - Brazil

^[26] Univ Sao Paulo, Inst Astron Geofis & Ciencias Atmosfer, Dept Astron, BR-05508090 Sao Paulo, SP - Brazil

^[27] Univ Alabama, Dept Phys & Astron, Box 870324, Tuscaloosa, AL 35487 - USA

^[28] Univ Michigan, Dept Astron, 311 West Hall, 1085 South Univ Ave, Ann Arbor, MI 48109 - USA

^[29] Instruments4, 4121 Pembury Pl, La Canada Flintridge, CA 91011 - USA

• Document type: Journal article
• Source: Astronomy & Astrophysics; v. 645, JAN 18 2021.
• Web of Science Citations: 0

Abstract

Context. Future astrophysical surveys such as J-PAS will produce very large datasets, the so-called ``big data{''}, which will require the deployment of accurate and efficient machine-learning (ML) methods. In this work, we analyze the miniJPAS survey, which observed about similar to 1 deg(2) of the AEGIS field with 56 narrow-band filters and 4 ugri broad-band filters. The miniJPAS primary catalog contains approximately 64 000 objects in the r detection band (mag(AB)less than or similar to 24), with forced-photometry in all other filters.Aims. We discuss the classification of miniJPAS sources into extended (galaxies) and point-like (e.g., stars) objects, which is a step required for the subsequent scientific analyses. We aim at developing an ML classifier that is complementary to traditional tools that are based on explicit modeling. In particular, our goal is to release a value-added catalog with our best classification.Methods. In order to train and test our classifiers, we cross-matched the miniJPAS dataset with SDSS and HSC-SSP data, whose classification is trustworthy within the intervals 15 <= r <= 20 and 18.5 <= r <= 23.5, respectively. We trained and tested six different ML algorithms on the two cross-matched catalogs: K-nearest neighbors, decision trees, random forest (RF), artificial neural networks, extremely randomized trees (ERT), and an ensemble classifier. This last is a hybrid algorithm that combines artificial neural networks and RF with the J-PAS stellar and galactic loci classifier. As input for the ML algorithms we used the magnitudes from the 60 filters together with their errors, with and without the morphological parameters. We also used the mean point spread function in the r detection band for each pointing.Results. We find that the RF and ERT algorithms perform best in all scenarios. When the full magnitude range of 15 <= r <= 23.5 is analyzed, we find an area under the curve AUC=0.957 with RF when photometric information alone is used, and AUC=0.986 with ERT when photometric and morphological information is used together. When morphological parameters are used, the full width at half maximum is the most important feature. When photometric information is used alone, we observe that broad bands are not necessarily more important than narrow bands, and errors (the width of the distribution) are as important as the measurements (central value of the distribution). In other words, it is apparently important to fully characterize the measurement.Conclusions. ML algorithms can compete with traditional star and galaxy classifiers; they outperform the latter at fainter magnitudes (r greater than or similar to 21). We use our best classifiers, with and without morphology, in order to produce a value-added catalog. (AU)

• FAPESP's process: 12/00800-4 - The 3-D universe: astrophysics with large galaxy surveys
• Grantee: Laerte Sodré Junior
• Support Opportunities: Research Projects - Thematic Grants