Reconstruction of decays to merged photons using end-to-end deep learning with domain continuation in the CMS detector
Autor(a) principal: | |
---|---|
Data de Publicação: | 2023 |
Outros Autores: | , , |
Tipo de documento: | Artigo |
Idioma: | eng |
Título da fonte: | Repositório Institucional da UFRGS |
Texto Completo: | http://hdl.handle.net/10183/272217 |
Resumo: | A novel technique based on machine learning is introduced to reconstruct the decays of highly Lorentz boosted particles. Using an end-to-end deep learning strategy, the technique bypasses existing rule-based particle reconstruction methods typically used in high energy physics analyses. It uses minimally processed detector data as input and directly outputs particle properties of interest. The new technique is demonstrated for the reconstruction of the invariant mass of particles decaying in the CMS detector. The decay of a hypothetical scalar particle A into two photons, A → γγ, is chosen as a benchmark decay. Lorentz boosts γL ¼ 60–600 are considered, ranging from regimes where both photons are resolved to those where the photons are closely merged as one object. A training method using domain continuation is introduced, enabling the invariant mass reconstruction of unresolved photon pairs in a novel way. The new technique is validated using π0 → γγ decays in LHC collision data. |
id |
UFRGS-2_c001f39f8297fbfc40c968ddd7a175a2 |
---|---|
oai_identifier_str |
oai:www.lume.ufrgs.br:10183/272217 |
network_acronym_str |
UFRGS-2 |
network_name_str |
Repositório Institucional da UFRGS |
repository_id_str |
|
spelling |
Tumasyan, ArmenSilveira, Gustavo Gil daBernardes, César AugustoCMS Collaboration2024-02-27T04:58:04Z20232470-0010http://hdl.handle.net/10183/272217001195248A novel technique based on machine learning is introduced to reconstruct the decays of highly Lorentz boosted particles. Using an end-to-end deep learning strategy, the technique bypasses existing rule-based particle reconstruction methods typically used in high energy physics analyses. It uses minimally processed detector data as input and directly outputs particle properties of interest. The new technique is demonstrated for the reconstruction of the invariant mass of particles decaying in the CMS detector. The decay of a hypothetical scalar particle A into two photons, A → γγ, is chosen as a benchmark decay. Lorentz boosts γL ¼ 60–600 are considered, ranging from regimes where both photons are resolved to those where the photons are closely merged as one object. A training method using domain continuation is introduced, enabling the invariant mass reconstruction of unresolved photon pairs in a novel way. The new technique is validated using π0 → γγ decays in LHC collision data.application/pdfengPhysical review. D, Particles, fields, gravitation, and cosmology. College Park. Vol. 108, no. 5 (Sept. 2023), 052002, 34 p.Aceleradores de partículasFotonsAprendizado de máquinaReconstruction of decays to merged photons using end-to-end deep learning with domain continuation in the CMS detectorEstrangeiroinfo:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/openAccessreponame:Repositório Institucional da UFRGSinstname:Universidade Federal do Rio Grande do Sul (UFRGS)instacron:UFRGSTEXT001195248.pdf.txt001195248.pdf.txtExtracted Texttext/plain154337http://www.lume.ufrgs.br/bitstream/10183/272217/2/001195248.pdf.txt5aa3ed634d8643e4bd7d0bc5f6459997MD52ORIGINAL001195248.pdfTexto completo (inglês)application/pdf37720869http://www.lume.ufrgs.br/bitstream/10183/272217/1/001195248.pdf8ff60bb239f3910da84b8363ee47903aMD5110183/2722172024-02-28 05:03:12.63652oai:www.lume.ufrgs.br:10183/272217Repositório de PublicaçõesPUBhttps://lume.ufrgs.br/oai/requestopendoar:2024-02-28T08:03:12Repositório Institucional da UFRGS - Universidade Federal do Rio Grande do Sul (UFRGS)false |
dc.title.pt_BR.fl_str_mv |
Reconstruction of decays to merged photons using end-to-end deep learning with domain continuation in the CMS detector |
title |
Reconstruction of decays to merged photons using end-to-end deep learning with domain continuation in the CMS detector |
spellingShingle |
Reconstruction of decays to merged photons using end-to-end deep learning with domain continuation in the CMS detector Tumasyan, Armen Aceleradores de partículas Fotons Aprendizado de máquina |
title_short |
Reconstruction of decays to merged photons using end-to-end deep learning with domain continuation in the CMS detector |
title_full |
Reconstruction of decays to merged photons using end-to-end deep learning with domain continuation in the CMS detector |
title_fullStr |
Reconstruction of decays to merged photons using end-to-end deep learning with domain continuation in the CMS detector |
title_full_unstemmed |
Reconstruction of decays to merged photons using end-to-end deep learning with domain continuation in the CMS detector |
title_sort |
Reconstruction of decays to merged photons using end-to-end deep learning with domain continuation in the CMS detector |
author |
Tumasyan, Armen |
author_facet |
Tumasyan, Armen Silveira, Gustavo Gil da Bernardes, César Augusto CMS Collaboration |
author_role |
author |
author2 |
Silveira, Gustavo Gil da Bernardes, César Augusto CMS Collaboration |
author2_role |
author author author |
dc.contributor.author.fl_str_mv |
Tumasyan, Armen Silveira, Gustavo Gil da Bernardes, César Augusto CMS Collaboration |
dc.subject.por.fl_str_mv |
Aceleradores de partículas Fotons Aprendizado de máquina |
topic |
Aceleradores de partículas Fotons Aprendizado de máquina |
description |
A novel technique based on machine learning is introduced to reconstruct the decays of highly Lorentz boosted particles. Using an end-to-end deep learning strategy, the technique bypasses existing rule-based particle reconstruction methods typically used in high energy physics analyses. It uses minimally processed detector data as input and directly outputs particle properties of interest. The new technique is demonstrated for the reconstruction of the invariant mass of particles decaying in the CMS detector. The decay of a hypothetical scalar particle A into two photons, A → γγ, is chosen as a benchmark decay. Lorentz boosts γL ¼ 60–600 are considered, ranging from regimes where both photons are resolved to those where the photons are closely merged as one object. A training method using domain continuation is introduced, enabling the invariant mass reconstruction of unresolved photon pairs in a novel way. The new technique is validated using π0 → γγ decays in LHC collision data. |
publishDate |
2023 |
dc.date.issued.fl_str_mv |
2023 |
dc.date.accessioned.fl_str_mv |
2024-02-27T04:58:04Z |
dc.type.driver.fl_str_mv |
Estrangeiro info:eu-repo/semantics/article |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
format |
article |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://hdl.handle.net/10183/272217 |
dc.identifier.issn.pt_BR.fl_str_mv |
2470-0010 |
dc.identifier.nrb.pt_BR.fl_str_mv |
001195248 |
identifier_str_mv |
2470-0010 001195248 |
url |
http://hdl.handle.net/10183/272217 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.relation.ispartof.pt_BR.fl_str_mv |
Physical review. D, Particles, fields, gravitation, and cosmology. College Park. Vol. 108, no. 5 (Sept. 2023), 052002, 34 p. |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
application/pdf |
dc.source.none.fl_str_mv |
reponame:Repositório Institucional da UFRGS instname:Universidade Federal do Rio Grande do Sul (UFRGS) instacron:UFRGS |
instname_str |
Universidade Federal do Rio Grande do Sul (UFRGS) |
instacron_str |
UFRGS |
institution |
UFRGS |
reponame_str |
Repositório Institucional da UFRGS |
collection |
Repositório Institucional da UFRGS |
bitstream.url.fl_str_mv |
http://www.lume.ufrgs.br/bitstream/10183/272217/2/001195248.pdf.txt http://www.lume.ufrgs.br/bitstream/10183/272217/1/001195248.pdf |
bitstream.checksum.fl_str_mv |
5aa3ed634d8643e4bd7d0bc5f6459997 8ff60bb239f3910da84b8363ee47903a |
bitstream.checksumAlgorithm.fl_str_mv |
MD5 MD5 |
repository.name.fl_str_mv |
Repositório Institucional da UFRGS - Universidade Federal do Rio Grande do Sul (UFRGS) |
repository.mail.fl_str_mv |
|
_version_ |
1801225111767875584 |