Particle Cloud Generation with Message Passing Generative Adversarial Networks
Autor(a) principal: | |
---|---|
Data de Publicação: | 2021 |
Outros Autores: | , , , , , , , |
Tipo de documento: | Artigo de conferência |
Idioma: | eng |
Título da fonte: | Repositório Institucional da UNESP |
Texto Completo: | http://hdl.handle.net/11449/240249 |
Resumo: | In high energy physics (HEP), jets are collections of correlated particles produced ubiquitously in particle collisions such as those at the CERN Large Hadron Collider (LHC). Machine learning (ML)-based generative models, such as generative adversarial networks (GANs), have the potential to significantly accelerate LHC jet simulations. However, despite jets having a natural representation as a set of particles in momentum-space, a.k.a. a particle cloud, there exist no generative models applied to such a dataset. In this work, we introduce a new particle cloud dataset (JetNet), and apply to it existing point cloud GANs. Results are evaluated using (1) 1-Wasserstein distances between high- and low-level feature distributions, (2) a newly developed Fréchet ParticleNet Distance, and (3) the coverage and (4) minimum matching distance metrics. Existing GANs are found to be inadequate for physics applications, hence we develop a new message passing GAN (MPGAN), which outperforms existing point cloud GANs on virtually every metric and shows promise for use in HEP. We propose JetNet as a novel point-cloud-style dataset for the ML community to experiment with, and set MPGAN as a benchmark to improve upon for future generative models. Additionally, to facilitate research and improve accessibility and reproducibility in this area, we release the open-source JETNET Python package with interfaces for particle cloud datasets, implementations for evaluation and loss metrics, and more tools for ML in HEP development. |
id |
UNSP_83cf7f6c28242602e50711a62862f5e7 |
---|---|
oai_identifier_str |
oai:repositorio.unesp.br:11449/240249 |
network_acronym_str |
UNSP |
network_name_str |
Repositório Institucional da UNESP |
repository_id_str |
2946 |
spelling |
Particle Cloud Generation with Message Passing Generative Adversarial NetworksIn high energy physics (HEP), jets are collections of correlated particles produced ubiquitously in particle collisions such as those at the CERN Large Hadron Collider (LHC). Machine learning (ML)-based generative models, such as generative adversarial networks (GANs), have the potential to significantly accelerate LHC jet simulations. However, despite jets having a natural representation as a set of particles in momentum-space, a.k.a. a particle cloud, there exist no generative models applied to such a dataset. In this work, we introduce a new particle cloud dataset (JetNet), and apply to it existing point cloud GANs. Results are evaluated using (1) 1-Wasserstein distances between high- and low-level feature distributions, (2) a newly developed Fréchet ParticleNet Distance, and (3) the coverage and (4) minimum matching distance metrics. Existing GANs are found to be inadequate for physics applications, hence we develop a new message passing GAN (MPGAN), which outperforms existing point cloud GANs on virtually every metric and shows promise for use in HEP. We propose JetNet as a novel point-cloud-style dataset for the ML community to experiment with, and set MPGAN as a benchmark to improve upon for future generative models. Additionally, to facilitate research and improve accessibility and reproducibility in this area, we release the open-source JETNET Python package with interfaces for particle cloud datasets, implementations for evaluation and loss metrics, and more tools for ML in HEP development.University of California San Diego, La JollaUniversidade Estadual Paulista, SPEuropean Organization for Nuclear Research (CERN)California Institute of TechnologyNational and Kapodistrian University of AthensUniversidade Estadual Paulista, SPUniversity of California San DiegoUniversidade Estadual Paulista (UNESP)European Organization for Nuclear Research (CERN)California Institute of TechnologyUniversity of AthensKansal, RaghavDuarte, JavierSu, HaoOrzari, Breno [UNESP]Tomei, Thiago [UNESP]Pierini, MaurizioTouranakou, MaryVlimant, Jean-RochGunopulos, Dimitrios2023-03-01T20:08:24Z2023-03-01T20:08:24Z2021-01-01info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/conferenceObject23858-23871Advances in Neural Information Processing Systems, v. 29, p. 23858-23871.1049-5258http://hdl.handle.net/11449/2402492-s2.0-85131957529Scopusreponame:Repositório Institucional da UNESPinstname:Universidade Estadual Paulista (UNESP)instacron:UNESPengAdvances in Neural Information Processing Systemsinfo:eu-repo/semantics/openAccess2023-03-01T20:08:24Zoai:repositorio.unesp.br:11449/240249Repositório InstitucionalPUBhttp://repositorio.unesp.br/oai/requestopendoar:29462023-03-01T20:08:24Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP)false |
dc.title.none.fl_str_mv |
Particle Cloud Generation with Message Passing Generative Adversarial Networks |
title |
Particle Cloud Generation with Message Passing Generative Adversarial Networks |
spellingShingle |
Particle Cloud Generation with Message Passing Generative Adversarial Networks Kansal, Raghav |
title_short |
Particle Cloud Generation with Message Passing Generative Adversarial Networks |
title_full |
Particle Cloud Generation with Message Passing Generative Adversarial Networks |
title_fullStr |
Particle Cloud Generation with Message Passing Generative Adversarial Networks |
title_full_unstemmed |
Particle Cloud Generation with Message Passing Generative Adversarial Networks |
title_sort |
Particle Cloud Generation with Message Passing Generative Adversarial Networks |
author |
Kansal, Raghav |
author_facet |
Kansal, Raghav Duarte, Javier Su, Hao Orzari, Breno [UNESP] Tomei, Thiago [UNESP] Pierini, Maurizio Touranakou, Mary Vlimant, Jean-Roch Gunopulos, Dimitrios |
author_role |
author |
author2 |
Duarte, Javier Su, Hao Orzari, Breno [UNESP] Tomei, Thiago [UNESP] Pierini, Maurizio Touranakou, Mary Vlimant, Jean-Roch Gunopulos, Dimitrios |
author2_role |
author author author author author author author author |
dc.contributor.none.fl_str_mv |
University of California San Diego Universidade Estadual Paulista (UNESP) European Organization for Nuclear Research (CERN) California Institute of Technology University of Athens |
dc.contributor.author.fl_str_mv |
Kansal, Raghav Duarte, Javier Su, Hao Orzari, Breno [UNESP] Tomei, Thiago [UNESP] Pierini, Maurizio Touranakou, Mary Vlimant, Jean-Roch Gunopulos, Dimitrios |
description |
In high energy physics (HEP), jets are collections of correlated particles produced ubiquitously in particle collisions such as those at the CERN Large Hadron Collider (LHC). Machine learning (ML)-based generative models, such as generative adversarial networks (GANs), have the potential to significantly accelerate LHC jet simulations. However, despite jets having a natural representation as a set of particles in momentum-space, a.k.a. a particle cloud, there exist no generative models applied to such a dataset. In this work, we introduce a new particle cloud dataset (JetNet), and apply to it existing point cloud GANs. Results are evaluated using (1) 1-Wasserstein distances between high- and low-level feature distributions, (2) a newly developed Fréchet ParticleNet Distance, and (3) the coverage and (4) minimum matching distance metrics. Existing GANs are found to be inadequate for physics applications, hence we develop a new message passing GAN (MPGAN), which outperforms existing point cloud GANs on virtually every metric and shows promise for use in HEP. We propose JetNet as a novel point-cloud-style dataset for the ML community to experiment with, and set MPGAN as a benchmark to improve upon for future generative models. Additionally, to facilitate research and improve accessibility and reproducibility in this area, we release the open-source JETNET Python package with interfaces for particle cloud datasets, implementations for evaluation and loss metrics, and more tools for ML in HEP development. |
publishDate |
2021 |
dc.date.none.fl_str_mv |
2021-01-01 2023-03-01T20:08:24Z 2023-03-01T20:08:24Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/conferenceObject |
format |
conferenceObject |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
Advances in Neural Information Processing Systems, v. 29, p. 23858-23871. 1049-5258 http://hdl.handle.net/11449/240249 2-s2.0-85131957529 |
identifier_str_mv |
Advances in Neural Information Processing Systems, v. 29, p. 23858-23871. 1049-5258 2-s2.0-85131957529 |
url |
http://hdl.handle.net/11449/240249 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.relation.none.fl_str_mv |
Advances in Neural Information Processing Systems |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
23858-23871 |
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_ |
1803046645446737920 |