NAOC Open IR
Robust Gaussian process regression based on iterative trimming
Li, Zhao-Zhou1; Li, Lu2,3; Shao, Zhengyi2,4
2021-07-01
Source PublicationASTRONOMY AND COMPUTING
ISSN2213-1337
Volume36Pages:10
AbstractThe Gaussian process (GP) regression can be severely biased when the data are contaminated by outliers. This paper presents a new robust GP regression algorithm that iteratively trims the most extreme data points. While the new algorithm retains the attractive properties of the standard GP as a nonparametric and flexible regression method, it can greatly improve the model accuracy for contaminated data even in the presence of extreme or abundant outliers. It is also easier to implement compared with previous robust GP variants that rely on approximate inference. Applied to a wide range of experiments with different contamination levels, the proposed method significantly outperforms the standard GP and the popular robust GP variant with the Student-t likelihood in most test cases. In addition, as a practical example in the astrophysical study, we show that this method can precisely determine the main-sequence ridge line in the color-magnitude diagram of star clusters. (C) 2021 Elsevier B.V. All rights reserved.
KeywordGaussian process Robust regression Outlier detection Ridge line Star clusters
Funding OrganizationNational Key Basic R&D Pro-gram of China ; National Key Basic R&D Pro-gram of China ; NSFC, China ; NSFC, China ; 111 project, China ; 111 project, China ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; National Key Basic R&D Pro-gram of China ; National Key Basic R&D Pro-gram of China ; NSFC, China ; NSFC, China ; 111 project, China ; 111 project, China ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; National Key Basic R&D Pro-gram of China ; National Key Basic R&D Pro-gram of China ; NSFC, China ; NSFC, China ; 111 project, China ; 111 project, China ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; National Key Basic R&D Pro-gram of China ; National Key Basic R&D Pro-gram of China ; NSFC, China ; NSFC, China ; 111 project, China ; 111 project, China ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education
DOI10.1016/j.ascom.2021.100483
WOS KeywordEFFICIENCY ; LTS
Language英语
Funding ProjectNational Key Basic R&D Pro-gram of China[2018YFA0404504] ; National Key Basic R&D Pro-gram of China[2019YFA0405501] ; NSFC, China[11621303] ; NSFC, China[11873038] ; NSFC, China[11890691] ; NSFC, China[11973032] ; NSFC, China[12022307] ; NSFC, China[U2031139] ; 111 project, China[B20019] ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education
Funding OrganizationNational Key Basic R&D Pro-gram of China ; National Key Basic R&D Pro-gram of China ; NSFC, China ; NSFC, China ; 111 project, China ; 111 project, China ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; National Key Basic R&D Pro-gram of China ; National Key Basic R&D Pro-gram of China ; NSFC, China ; NSFC, China ; 111 project, China ; 111 project, China ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; National Key Basic R&D Pro-gram of China ; National Key Basic R&D Pro-gram of China ; NSFC, China ; NSFC, China ; 111 project, China ; 111 project, China ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; National Key Basic R&D Pro-gram of China ; National Key Basic R&D Pro-gram of China ; NSFC, China ; NSFC, China ; 111 project, China ; 111 project, China ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education ; MOE Key Lab for Particle Physics, Astrophysics and Cosmology, Ministry of Education
WOS Research AreaAstronomy & Astrophysics ; Computer Science
WOS SubjectAstronomy & Astrophysics ; Computer Science, Interdisciplinary Applications
WOS IDWOS:000691531100010
PublisherELSEVIER
Citation statistics
Document Type期刊论文
Identifierhttp://ir.bao.ac.cn/handle/114a11/75380
Collection中国科学院国家天文台
Corresponding AuthorLi, Zhao-Zhou
Affiliation1.Shanghai Jiao Tong Univ, Sch Phys & Astron, Dept Astron, 955 Jianchuan Rd, Shanghai 200240, Peoples R China
2.Chinese Acad Sci, Shanghai Astron Observ, Key Lab Res Galaxies & Cosmol, 80 Nandan Rd, Shanghai 200030, Peoples R China
3.Univ Chinese Acad Sci, 19A Yuquan Rd, Beijing 100049, Peoples R China
4.Key Lab Astrophys, Shanghai 200234, Peoples R China
Recommended Citation
GB/T 7714
Li, Zhao-Zhou,Li, Lu,Shao, Zhengyi. Robust Gaussian process regression based on iterative trimming[J]. ASTRONOMY AND COMPUTING,2021,36:10.
APA Li, Zhao-Zhou,Li, Lu,&Shao, Zhengyi.(2021).Robust Gaussian process regression based on iterative trimming.ASTRONOMY AND COMPUTING,36,10.
MLA Li, Zhao-Zhou,et al."Robust Gaussian process regression based on iterative trimming".ASTRONOMY AND COMPUTING 36(2021):10.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Li, Zhao-Zhou]'s Articles
[Li, Lu]'s Articles
[Shao, Zhengyi]'s Articles
Baidu academic
Similar articles in Baidu academic
[Li, Zhao-Zhou]'s Articles
[Li, Lu]'s Articles
[Shao, Zhengyi]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Li, Zhao-Zhou]'s Articles
[Li, Lu]'s Articles
[Shao, Zhengyi]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.