Cargando…
Improving Variational Autoencoders for New Physics Detection at the LHC with Normalizing Flows
We investigate how to improve new physics detection strategies exploiting variational autoencoders and normalizing flows for anomaly detection at the Large Hadron Collider. As a working example, we consider the DarkMachines challenge dataset. We show how different design choices (e.g., event represe...
| Autores principales: | , , , , , , , |
|---|---|
| Lenguaje: | eng |
| Publicado: |
2021
|
| Materias: | |
| Acceso en línea: | https://dx.doi.org/10.3389/fdata.2022.803685 http://cds.cern.ch/record/2784906 |
| _version_ | 1780972111536324608 |
|---|---|
| author | Jawahar, Pratik Aarrestad, Thea Chernyavskaya, Nadezda Pierini, Maurizio Wozniak, Kinga A. Ngadiuba, Jennifer Duarte, Javier Tsan, Steven |
| author_facet | Jawahar, Pratik Aarrestad, Thea Chernyavskaya, Nadezda Pierini, Maurizio Wozniak, Kinga A. Ngadiuba, Jennifer Duarte, Javier Tsan, Steven |
| author_sort | Jawahar, Pratik |
| collection | CERN |
| description | We investigate how to improve new physics detection strategies exploiting variational autoencoders and normalizing flows for anomaly detection at the Large Hadron Collider. As a working example, we consider the DarkMachines challenge dataset. We show how different design choices (e.g., event representations, anomaly score definitions, network architectures) affect the result on specific benchmark new physics models. Once a baseline is established, we discuss how to improve the anomaly detection accuracy by exploiting normalizing flow layers in the latent space of the variational autoencoder. |
| id | cern-2784906 |
| institution | Organización Europea para la Investigación Nuclear |
| language | eng |
| publishDate | 2021 |
| record_format | invenio |
| spelling | cern-27849062023-07-27T08:30:21Zdoi:10.3389/fdata.2022.803685http://cds.cern.ch/record/2784906engJawahar, PratikAarrestad, TheaChernyavskaya, NadezdaPierini, MaurizioWozniak, Kinga A.Ngadiuba, JenniferDuarte, JavierTsan, StevenImproving Variational Autoencoders for New Physics Detection at the LHC with Normalizing Flowsphysics.data-anOther Fields of Physicshep-exParticle Physics - Experimenthep-phParticle Physics - PhenomenologyWe investigate how to improve new physics detection strategies exploiting variational autoencoders and normalizing flows for anomaly detection at the Large Hadron Collider. As a working example, we consider the DarkMachines challenge dataset. We show how different design choices (e.g., event representations, anomaly score definitions, network architectures) affect the result on specific benchmark new physics models. Once a baseline is established, we discuss how to improve the anomaly detection accuracy by exploiting normalizing flow layers in the latent space of the variational autoencoder.We investigate how to improve new physics detection strategies exploiting variational autoencoders and normalizing flows for anomaly detection at the Large Hadron Collider. As a working example, we consider the DarkMachines challenge dataset. We show how different design choices (e.g., event representations, anomaly score definitions, network architectures) affect the result on specific benchmark new physics models. Once a baseline is established, we discuss how to improve the anomaly detection accuracy by exploiting normalizing flow layers in the latent space of the variational autoencoder.arXiv:2110.08508FERMILAB-PUB-21-519-CMSoai:cds.cern.ch:27849062021-10-16 |
| spellingShingle | physics.data-an Other Fields of Physics hep-ex Particle Physics - Experiment hep-ph Particle Physics - Phenomenology Jawahar, Pratik Aarrestad, Thea Chernyavskaya, Nadezda Pierini, Maurizio Wozniak, Kinga A. Ngadiuba, Jennifer Duarte, Javier Tsan, Steven Improving Variational Autoencoders for New Physics Detection at the LHC with Normalizing Flows |
| title | Improving Variational Autoencoders for New Physics Detection at the LHC with Normalizing Flows |
| title_full | Improving Variational Autoencoders for New Physics Detection at the LHC with Normalizing Flows |
| title_fullStr | Improving Variational Autoencoders for New Physics Detection at the LHC with Normalizing Flows |
| title_full_unstemmed | Improving Variational Autoencoders for New Physics Detection at the LHC with Normalizing Flows |
| title_short | Improving Variational Autoencoders for New Physics Detection at the LHC with Normalizing Flows |
| title_sort | improving variational autoencoders for new physics detection at the lhc with normalizing flows |
| topic | physics.data-an Other Fields of Physics hep-ex Particle Physics - Experiment hep-ph Particle Physics - Phenomenology |
| url | https://dx.doi.org/10.3389/fdata.2022.803685 http://cds.cern.ch/record/2784906 |
| work_keys_str_mv | AT jawaharpratik improvingvariationalautoencodersfornewphysicsdetectionatthelhcwithnormalizingflows AT aarrestadthea improvingvariationalautoencodersfornewphysicsdetectionatthelhcwithnormalizingflows AT chernyavskayanadezda improvingvariationalautoencodersfornewphysicsdetectionatthelhcwithnormalizingflows AT pierinimaurizio improvingvariationalautoencodersfornewphysicsdetectionatthelhcwithnormalizingflows AT wozniakkingaa improvingvariationalautoencodersfornewphysicsdetectionatthelhcwithnormalizingflows AT ngadiubajennifer improvingvariationalautoencodersfornewphysicsdetectionatthelhcwithnormalizingflows AT duartejavier improvingvariationalautoencodersfornewphysicsdetectionatthelhcwithnormalizingflows AT tsansteven improvingvariationalautoencodersfornewphysicsdetectionatthelhcwithnormalizingflows |