Particle Cloud Generation with Message Passing Generative Adversarial Networks

Registro completo de metadados
MetadadosDescriçãoIdioma
Autor(es): dc.contributorUniv Calif San Diego-
Autor(es): dc.contributorUniversidade Estadual Paulista (UNESP)-
Autor(es): dc.contributorEuropean Org Nucl Res (CERN-
Autor(es): dc.contributorCALTECH-
Autor(es): dc.contributorNatl & Kapodistrian Univ Athens-
Autor(es): dc.creatorKansal, Raghav-
Autor(es): dc.creatorDuarte, Javier-
Autor(es): dc.creatorSu, Hao-
Autor(es): dc.creatorOrzari, Breno-
Autor(es): dc.creatorTomei, Thiago-
Autor(es): dc.creatorPierini, Maurizio-
Autor(es): dc.creatorTouranakou, Mary-
Autor(es): dc.creatorVlimant, Jean-Roch-
Autor(es): dc.creatorRanzato, M.-
Autor(es): dc.creatorBeygelzimer, A.-
Autor(es): dc.creatorDauphin, Y.-
Autor(es): dc.creatorLiang, P. S.-
Autor(es): dc.creatorVaughan, J. W.-
Data de aceite: dc.date.accessioned2025-08-21T21:57:33Z-
Data de disponibilização: dc.date.available2025-08-21T21:57:33Z-
Data de envio: dc.date.issued2023-07-29-
Data de envio: dc.date.issued2023-07-29-
Data de envio: dc.date.issued2020-12-31-
Fonte completa do material: dc.identifierhttp://hdl.handle.net/11449/245186-
Fonte: dc.identifier.urihttp://educapes.capes.gov.br/handle/11449/245186-
Descrição: dc.descriptionIn 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 Frechet 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.-
Descrição: dc.descriptionEuropean Research Council (ERC) under the European Union-
Descrição: dc.descriptionIRIS-HEP fellowship through the U.S. National Science Foundation (NSF)-
Descrição: dc.descriptionU.S. Department of Energy (DOE)-
Descrição: dc.descriptionDOE, Office of Science, Office of High Energy Physics Early Career Research program-
Descrição: dc.descriptionDOE, Office of Advanced Scientific Computing Research-
Descrição: dc.descriptionFunda��o de Amparo � Pesquisa do Estado de S�o Paulo (FAPESP)-
Descrição: dc.descriptionERC under the European Union-
Descrição: dc.descriptionDOE, Office of Science, Office of High Energy Physics-
Descrição: dc.descriptionEU-
Descrição: dc.descriptionNSF-
Descrição: dc.descriptionUniversity of California Office of the President-
Descrição: dc.descriptionUniversity of California San Diego's California Institute for Telecommunications and Information Technology/Qualcomm Institute-
Descrição: dc.descriptionUniv Calif San Diego, La Jolla, CA 92093 USA-
Descrição: dc.descriptionUniv Estadual Paulista, BR-01049010 Sao Paulo, SP, Brazil-
Descrição: dc.descriptionEuropean Org Nucl Res (CERN, CH-1211 Geneva 23, Switzerland-
Descrição: dc.descriptionCALTECH, Pasadena, CA 91125 USA-
Descrição: dc.descriptionNatl & Kapodistrian Univ Athens, Athens, Greece-
Descrição: dc.descriptionUniv Estadual Paulista, BR-01049010 Sao Paulo, SP, Brazil-
Descrição: dc.descriptionEuropean Research Council (ERC) under the European Union: 772369-
Descrição: dc.descriptionIRIS-HEP fellowship through the U.S. National Science Foundation (NSF): OAC-1836650-
Descrição: dc.descriptionU.S. Department of Energy (DOE): DE-AC0207CH11359-
Descrição: dc.descriptionDOE, Office of Science, Office of High Energy Physics Early Career Research program: DESC0021187-
Descrição: dc.descriptionDOE, Office of Advanced Scientific Computing Research: DE-SC0021396-
Descrição: dc.descriptionFAPESP: 2018/25225-9-
Descrição: dc.descriptionFAPESP: 2018/01398-1-
Descrição: dc.descriptionFAPESP: 2019/16401-0-
Descrição: dc.descriptionERC under the European Union: 772369-
Descrição: dc.descriptionDOE, Office of Science, Office of High Energy Physics: DE-SC0011925-
Descrição: dc.descriptionDOE, Office of Science, Office of High Energy Physics: DE-SC0019227-
Descrição: dc.descriptionDOE, Office of Science, Office of High Energy Physics: DE-AC02-07CH11359-
Descrição: dc.descriptionEU: 952215-
Descrição: dc.descriptionNSF: 1904444-
Descrição: dc.descriptionNSF: CNS-1730158-
Descrição: dc.descriptionNSF: ACI1540112-
Descrição: dc.descriptionNSF: ACI-1541349-
Descrição: dc.descriptionNSF: OAC-1826967-
Formato: dc.format14-
Idioma: dc.languageen-
Publicador: dc.publisherNeural Information Processing Systems (nips)-
Relação: dc.relationAdvances In Neural Information Processing Systems 34 (neurips 2021)-
???dc.source???: dc.sourceWeb of Science-
Título: dc.titleParticle Cloud Generation with Message Passing Generative Adversarial Networks-
Tipo de arquivo: dc.typeaula digital-
Aparece nas coleções:Repositório Institucional - Unesp

Não existem arquivos associados a este item.