Optimization of artificial neural networks models applied to the identification of images of asteroids’ resonant arguments
Autor(a) principal: | |
---|---|
Data de Publicação: | 2022 |
Outros Autores: | , , , |
Tipo de documento: | Artigo |
Idioma: | eng |
Título da fonte: | Repositório Institucional da UNESP |
Texto Completo: | http://dx.doi.org/10.1007/s10569-022-10110-7 http://hdl.handle.net/11449/249487 |
Resumo: | The asteroidal main belt is crossed by a web of mean motion and secular resonances that occur when there is a commensurability between fundamental frequencies of the asteroids and planets. Traditionally, these objects were identified by visual inspection of the time evolution of their resonant argument, which is a combination of orbital elements of the asteroid and the perturbing planet(s). Since the population of asteroids affected by these resonances is, in some cases, of the order of several thousand, this has become a taxing task for a human observer. Recent works used convolutional neural network (CNN) models to perform such task automatically. In this work, we compare the outcome of such models with those of some of the most advanced and publicly available CNN architectures, like the VGG, Inception, and ResNet. The performance of such models is first tested and optimized for overfitting issues, using validation sets and a series of regularization techniques like data augmentation, dropout, and batch normalization. The three best-performing models were then used to predict the labels of larger testing databases containing thousands of images. The VGG model, with and without regularizations, proved to be the most efficient method to predict labels of large datasets. Since the Vera C. Rubin observatory is likely to discover up to four million new asteroids in the next few years, the use of these models might become quite valuable to identify populations of resonant minor bodies. |
id |
UNSP_88c9a0be9278187bff6b06ca6a942187 |
---|---|
oai_identifier_str |
oai:repositorio.unesp.br:11449/249487 |
network_acronym_str |
UNSP |
network_name_str |
Repositório Institucional da UNESP |
repository_id_str |
2946 |
spelling |
Optimization of artificial neural networks models applied to the identification of images of asteroids’ resonant argumentsAsteroidsGeneralMinor planetsTime domain astronomyTime series analysisThe asteroidal main belt is crossed by a web of mean motion and secular resonances that occur when there is a commensurability between fundamental frequencies of the asteroids and planets. Traditionally, these objects were identified by visual inspection of the time evolution of their resonant argument, which is a combination of orbital elements of the asteroid and the perturbing planet(s). Since the population of asteroids affected by these resonances is, in some cases, of the order of several thousand, this has become a taxing task for a human observer. Recent works used convolutional neural network (CNN) models to perform such task automatically. In this work, we compare the outcome of such models with those of some of the most advanced and publicly available CNN architectures, like the VGG, Inception, and ResNet. The performance of such models is first tested and optimized for overfitting issues, using validation sets and a series of regularization techniques like data augmentation, dropout, and batch normalization. The three best-performing models were then used to predict the labels of larger testing databases containing thousands of images. The VGG model, with and without regularizations, proved to be the most efficient method to predict labels of large datasets. Since the Vera C. Rubin observatory is likely to discover up to four million new asteroids in the next few years, the use of these models might become quite valuable to identify populations of resonant minor bodies.Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)School of Natural Sciences and Engineering São Paulo State University (UNESP), SPDivision of Space Mechanics and Control National Space Research Institute (INPE), C.P. 515, SPSão Paulo State University (UNESP), SPSchool of Natural Sciences and Engineering São Paulo State University (UNESP), SPSão Paulo State University (UNESP), SPFAPESP: 2016/024561-0CNPq: 304168/2021-1CAPES: 88887.675709/2022-00Universidade Estadual Paulista (UNESP)National Space Research Institute (INPE)Carruba, V. [UNESP]Aljbaae, S.Caritá, G.Domingos, R. C. [UNESP]Martins, B. [UNESP]2023-07-29T16:00:58Z2023-07-29T16:00:58Z2022-12-01info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/articlehttp://dx.doi.org/10.1007/s10569-022-10110-7Celestial Mechanics and Dynamical Astronomy, v. 134, n. 6, 2022.1572-94780923-2958http://hdl.handle.net/11449/24948710.1007/s10569-022-10110-72-s2.0-85144320705Scopusreponame:Repositório Institucional da UNESPinstname:Universidade Estadual Paulista (UNESP)instacron:UNESPengCelestial Mechanics and Dynamical Astronomyinfo:eu-repo/semantics/openAccess2024-07-02T14:28:55Zoai:repositorio.unesp.br:11449/249487Repositório InstitucionalPUBhttp://repositorio.unesp.br/oai/requestopendoar:29462024-08-05T16:24:04.180598Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP)false |
dc.title.none.fl_str_mv |
Optimization of artificial neural networks models applied to the identification of images of asteroids’ resonant arguments |
title |
Optimization of artificial neural networks models applied to the identification of images of asteroids’ resonant arguments |
spellingShingle |
Optimization of artificial neural networks models applied to the identification of images of asteroids’ resonant arguments Carruba, V. [UNESP] Asteroids General Minor planets Time domain astronomy Time series analysis |
title_short |
Optimization of artificial neural networks models applied to the identification of images of asteroids’ resonant arguments |
title_full |
Optimization of artificial neural networks models applied to the identification of images of asteroids’ resonant arguments |
title_fullStr |
Optimization of artificial neural networks models applied to the identification of images of asteroids’ resonant arguments |
title_full_unstemmed |
Optimization of artificial neural networks models applied to the identification of images of asteroids’ resonant arguments |
title_sort |
Optimization of artificial neural networks models applied to the identification of images of asteroids’ resonant arguments |
author |
Carruba, V. [UNESP] |
author_facet |
Carruba, V. [UNESP] Aljbaae, S. Caritá, G. Domingos, R. C. [UNESP] Martins, B. [UNESP] |
author_role |
author |
author2 |
Aljbaae, S. Caritá, G. Domingos, R. C. [UNESP] Martins, B. [UNESP] |
author2_role |
author author author author |
dc.contributor.none.fl_str_mv |
Universidade Estadual Paulista (UNESP) National Space Research Institute (INPE) |
dc.contributor.author.fl_str_mv |
Carruba, V. [UNESP] Aljbaae, S. Caritá, G. Domingos, R. C. [UNESP] Martins, B. [UNESP] |
dc.subject.por.fl_str_mv |
Asteroids General Minor planets Time domain astronomy Time series analysis |
topic |
Asteroids General Minor planets Time domain astronomy Time series analysis |
description |
The asteroidal main belt is crossed by a web of mean motion and secular resonances that occur when there is a commensurability between fundamental frequencies of the asteroids and planets. Traditionally, these objects were identified by visual inspection of the time evolution of their resonant argument, which is a combination of orbital elements of the asteroid and the perturbing planet(s). Since the population of asteroids affected by these resonances is, in some cases, of the order of several thousand, this has become a taxing task for a human observer. Recent works used convolutional neural network (CNN) models to perform such task automatically. In this work, we compare the outcome of such models with those of some of the most advanced and publicly available CNN architectures, like the VGG, Inception, and ResNet. The performance of such models is first tested and optimized for overfitting issues, using validation sets and a series of regularization techniques like data augmentation, dropout, and batch normalization. The three best-performing models were then used to predict the labels of larger testing databases containing thousands of images. The VGG model, with and without regularizations, proved to be the most efficient method to predict labels of large datasets. Since the Vera C. Rubin observatory is likely to discover up to four million new asteroids in the next few years, the use of these models might become quite valuable to identify populations of resonant minor bodies. |
publishDate |
2022 |
dc.date.none.fl_str_mv |
2022-12-01 2023-07-29T16:00:58Z 2023-07-29T16:00:58Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/article |
format |
article |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://dx.doi.org/10.1007/s10569-022-10110-7 Celestial Mechanics and Dynamical Astronomy, v. 134, n. 6, 2022. 1572-9478 0923-2958 http://hdl.handle.net/11449/249487 10.1007/s10569-022-10110-7 2-s2.0-85144320705 |
url |
http://dx.doi.org/10.1007/s10569-022-10110-7 http://hdl.handle.net/11449/249487 |
identifier_str_mv |
Celestial Mechanics and Dynamical Astronomy, v. 134, n. 6, 2022. 1572-9478 0923-2958 10.1007/s10569-022-10110-7 2-s2.0-85144320705 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.relation.none.fl_str_mv |
Celestial Mechanics and Dynamical Astronomy |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.source.none.fl_str_mv |
Scopus reponame:Repositório Institucional da UNESP instname:Universidade Estadual Paulista (UNESP) instacron:UNESP |
instname_str |
Universidade Estadual Paulista (UNESP) |
instacron_str |
UNESP |
institution |
UNESP |
reponame_str |
Repositório Institucional da UNESP |
collection |
Repositório Institucional da UNESP |
repository.name.fl_str_mv |
Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP) |
repository.mail.fl_str_mv |
|
_version_ |
1808128644100390912 |