Machine learning for searching the dark energy survey for trans-neptunian objects

DES Collaboration

doi:10.1088/1538-3873/abcaea

Machine learning for searching the dark energy survey for trans-neptunian objects

DES Collaboration

Workflow Systems

Research output: Contribution to journal › Article › peer-review

7 Scopus citations

Abstract

In this paper we investigate how implementing machine learning could improve the efficiency of the search for Trans-Neptunian Objects (TNOs) within Dark Energy Survey (DES) data when used alongside orbit fitting. The discovery of multiple TNOs that appear to show a similarity in their orbital parameters has led to the suggestion that one or more undetected planets, an as yet undiscovered “Planet 9”, may be present in the outer solar system. DES is well placed to detect such a planet and has already been used to discover many other TNOs. Here, we perform tests on eight different supervised machine learning algorithms, using a data set consisting of simulated TNOs buried within real DES noise data. We found that the best performing classifier was the Random Forest which, when optimized, performed well at detecting the rare objects. We achieve an area under the receiver operating characteristic (ROC) curve, (AUC)=0.996±0.001. After optimizing the decision threshold of the Random Forest, we achieve a recall of 0.96 while maintaining a precision of 0.80. Finally, by using the optimized classifier to pre-select objects, we are able to run the orbit-fitting stage of our detection pipeline five times faster.

Original language	English
Article number	014501
Pages (from-to)	1-14
Number of pages	14
Journal	Publications of the Astronomical Society of the Pacific
Volume	133
Issue number	1019
DOIs	https://doi.org/10.1088/1538-3873/abcaea
State	Published - Jan 2021

Funding

Funding for the DES Projects has been provided by the U.S. Department of Energy, the U.S. National Science Foundation, the Ministry of Science and Education of Spain, the Science and Technology Facilities Council of the United Kingdom, the Higher Education Funding Council for England, the National Center for Supercomputing Applications at the University of Illinois at Urbana-Champaign, the Kavli Institute of Cosmological Physics at the University of Chicago, the Center for Cosmology and Astro-Particle Physics at the Ohio State University, the Mitchell Institute for Fundamental Physics and Astronomy at Texas A&M University, Financiadora de Estudos e Projetos, Funda\u00E7\u00E3o Carlos Chagas Filho de Amparo \u00E0 Pesquisa do Estado do Rio de Janeiro, Conselho Nacional de Desenvolvimento Cient\u00EDfico e Tecnol\u00F3gico and the Minist\u00E9rio da Ci\u00EAncia, Tecnologia e Inova\u00E7\u00E3o, the Deutsche Forschungs-gemeinschaft and the Collaborating Institutions in the Dark Energy Survey. The DES data management system is supported by the National Science Foundation under grant Nos. AST-1138766 and AST-1536171. The DES participants from Spanish institutions are partially supported by MICINN under grants ESP2017-89838, PGC2018-094773, PGC2018-102021, SEV-2016-0588, SEV-2016-0597, and MDM-2015-0509, some of which include ERDF funds from the European Union. IFAE is partially funded by the CERCA program of the Generalitat de Catalunya. Research leading to these results has received funding from the European Research Council under the European Union\u2019s Seventh Framework Program (FP7/2007-2013) including ERC grant agreements 240672, 291329, and 306478. We acknowledge support from the Brazilian Instituto Nacional de Ci\u00EAncia e Tecnologia (INCT) do e-Universo (CNPq grant 465376/2014-2). This manuscript has been authored by Fermi Research Alliance, LLC under Contract No. DE-AC02-07CH11359 with the U.S. Department of Energy, Office of Science, Office of High Energy Physics. B.H. was supported by the STFC UCL Centre for Doctoral Training in Data Intensive Science (grant No. ST/P006736/1). O.L. acknowledges support from a European Research Council Advanced Grant TESTDE FP7/291329 and an STFC Consolidated Grants ST/M001334/1 and ST/R000476/1.

Keywords

Computational methods
Minor planets
Random Forests
Trans-Neptunian objects

Access to Document

10.1088/1538-3873/abcaea

Cite this

@article{d3b18a59d9d145acba3213ffba6c3ea9,

title = "Machine learning for searching the dark energy survey for trans-neptunian objects",

abstract = "In this paper we investigate how implementing machine learning could improve the efficiency of the search for Trans-Neptunian Objects (TNOs) within Dark Energy Survey (DES) data when used alongside orbit fitting. The discovery of multiple TNOs that appear to show a similarity in their orbital parameters has led to the suggestion that one or more undetected planets, an as yet undiscovered “Planet 9”, may be present in the outer solar system. DES is well placed to detect such a planet and has already been used to discover many other TNOs. Here, we perform tests on eight different supervised machine learning algorithms, using a data set consisting of simulated TNOs buried within real DES noise data. We found that the best performing classifier was the Random Forest which, when optimized, performed well at detecting the rare objects. We achieve an area under the receiver operating characteristic (ROC) curve, (AUC)=0.996±0.001. After optimizing the decision threshold of the Random Forest, we achieve a recall of 0.96 while maintaining a precision of 0.80. Finally, by using the optimized classifier to pre-select objects, we are able to run the orbit-fitting stage of our detection pipeline five times faster.",

keywords = "Computational methods, Minor planets, Random Forests, Trans-Neptunian objects",

author = "{DES Collaboration} and B. Henghes and O. Lahav and Gerdes, {D. W.} and Lin, {H. W.} and R. Morgan and Abbott, {T. M.C.} and M. Aguena and S. Allam and J. Annis and S. Avila and E. Bertin and D. Brooks and Burke, {D. L.} and {Carnero Rosell}, A. and {Carrasco Kind}, M. and J. Carretero and C. Conselice and M. Costanzi and {da Costa}, {L. N.} and {De Vicente}, J. and S. Desai and Diehl, {H. T.} and P. Doel and S. Everett and I. Ferrero and J. Frieman and J. Garc{\'i}a-Bellido and E. Gaztanaga and D. Gruen and Gruendl, {R. A.} and J. Gschwend and G. Gutierrez and Hartley, {W. G.} and Hinton, {S. R.} and K. Honscheid and B. Hoyle and James, {D. J.} and K. Kuehn and N. Kuropatkin and Marshall, {J. L.} and P. Melchior and F. Menanteau and R. Miquel and Ogando, {R. L.C.} and A. Palmese and F. Paz-Chinch{\'o}n and Plazas, {A. A.} and Romer, {A. K.} and C. S{\'a}nchez and E. Suchyta",

year = "2021",

month = jan,

doi = "10.1088/1538-3873/abcaea",

language = "English",

volume = "133",

pages = "1--14",

journal = "Publications of the Astronomical Society of the Pacific",

issn = "0004-6280",

number = "1019",

}

TY - JOUR

T1 - Machine learning for searching the dark energy survey for trans-neptunian objects

AU - DES Collaboration

AU - Henghes, B.

AU - Lahav, O.

AU - Gerdes, D. W.

AU - Lin, H. W.

AU - Morgan, R.

AU - Abbott, T. M.C.

AU - Aguena, M.

AU - Allam, S.

AU - Annis, J.

AU - Avila, S.

AU - Bertin, E.

AU - Brooks, D.

AU - Burke, D. L.

AU - Carnero Rosell, A.

AU - Carrasco Kind, M.

AU - Carretero, J.

AU - Conselice, C.

AU - Costanzi, M.

AU - da Costa, L. N.

AU - De Vicente, J.

AU - Desai, S.

AU - Diehl, H. T.

AU - Doel, P.

AU - Everett, S.

AU - Ferrero, I.

AU - Frieman, J.

AU - García-Bellido, J.

AU - Gaztanaga, E.

AU - Gruen, D.

AU - Gruendl, R. A.

AU - Gschwend, J.

AU - Gutierrez, G.

AU - Hartley, W. G.

AU - Hinton, S. R.

AU - Honscheid, K.

AU - Hoyle, B.

AU - James, D. J.

AU - Kuehn, K.

AU - Kuropatkin, N.

AU - Marshall, J. L.

AU - Melchior, P.

AU - Menanteau, F.

AU - Miquel, R.

AU - Ogando, R. L.C.

AU - Palmese, A.

AU - Paz-Chinchón, F.

AU - Plazas, A. A.

AU - Romer, A. K.

AU - Sánchez, C.

AU - Suchyta, E.

PY - 2021/1

Y1 - 2021/1

N2 - In this paper we investigate how implementing machine learning could improve the efficiency of the search for Trans-Neptunian Objects (TNOs) within Dark Energy Survey (DES) data when used alongside orbit fitting. The discovery of multiple TNOs that appear to show a similarity in their orbital parameters has led to the suggestion that one or more undetected planets, an as yet undiscovered “Planet 9”, may be present in the outer solar system. DES is well placed to detect such a planet and has already been used to discover many other TNOs. Here, we perform tests on eight different supervised machine learning algorithms, using a data set consisting of simulated TNOs buried within real DES noise data. We found that the best performing classifier was the Random Forest which, when optimized, performed well at detecting the rare objects. We achieve an area under the receiver operating characteristic (ROC) curve, (AUC)=0.996±0.001. After optimizing the decision threshold of the Random Forest, we achieve a recall of 0.96 while maintaining a precision of 0.80. Finally, by using the optimized classifier to pre-select objects, we are able to run the orbit-fitting stage of our detection pipeline five times faster.

AB - In this paper we investigate how implementing machine learning could improve the efficiency of the search for Trans-Neptunian Objects (TNOs) within Dark Energy Survey (DES) data when used alongside orbit fitting. The discovery of multiple TNOs that appear to show a similarity in their orbital parameters has led to the suggestion that one or more undetected planets, an as yet undiscovered “Planet 9”, may be present in the outer solar system. DES is well placed to detect such a planet and has already been used to discover many other TNOs. Here, we perform tests on eight different supervised machine learning algorithms, using a data set consisting of simulated TNOs buried within real DES noise data. We found that the best performing classifier was the Random Forest which, when optimized, performed well at detecting the rare objects. We achieve an area under the receiver operating characteristic (ROC) curve, (AUC)=0.996±0.001. After optimizing the decision threshold of the Random Forest, we achieve a recall of 0.96 while maintaining a precision of 0.80. Finally, by using the optimized classifier to pre-select objects, we are able to run the orbit-fitting stage of our detection pipeline five times faster.

KW - Computational methods

KW - Minor planets

KW - Random Forests

KW - Trans-Neptunian objects

UR - http://www.scopus.com/inward/record.url?scp=85099796162&partnerID=8YFLogxK

U2 - 10.1088/1538-3873/abcaea

DO - 10.1088/1538-3873/abcaea

M3 - Article

AN - SCOPUS:85099796162

SN - 0004-6280

VL - 133

SP - 1

EP - 14

JO - Publications of the Astronomical Society of the Pacific

JF - Publications of the Astronomical Society of the Pacific

IS - 1019

M1 - 014501

ER -

Machine learning for searching the dark energy survey for trans-neptunian objects

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this