Cargando…

Robust Neural Particle Identification Models

The volume of data processed by the Large Hadron Collider experiments demands sophisticated selection rules typically based on machine learning algorithms. One of the shortcomings of these approaches is their profound sensitivity to the biases in training samples. In the case of particle identificat...

Descripción completa

Detalles Bibliográficos
Autores principales: Ryzhikov, Artem, Temirkhanov, Aziz, Derkach, Denis, Hushchyn, Mikhail, Kazeev, Nikita, Mokhnenko, Sergei
Lenguaje:eng
Publicado: 2023
Materias:
Acceso en línea:https://dx.doi.org/10.1088/1742-6596/2438/1/012119
http://cds.cern.ch/record/2850401
_version_ 1780977075402833920
author Ryzhikov, Artem
Temirkhanov, Aziz
Derkach, Denis
Hushchyn, Mikhail
Kazeev, Nikita
Mokhnenko, Sergei
author_facet Ryzhikov, Artem
Temirkhanov, Aziz
Derkach, Denis
Hushchyn, Mikhail
Kazeev, Nikita
Mokhnenko, Sergei
author_sort Ryzhikov, Artem
collection CERN
description The volume of data processed by the Large Hadron Collider experiments demands sophisticated selection rules typically based on machine learning algorithms. One of the shortcomings of these approaches is their profound sensitivity to the biases in training samples. In the case of particle identification (PID), this might lead to degradation of the efficiency for some decays not present in the training dataset due to differences in input kinematic distributions. In this talk, we propose a method based on the Common Specific Decomposition that takes into account individual decays and possible misshapes in the training data by disentangling common and decay specific components of the input feature set. We show that the proposed approach reduces the rate of efficiency degradation for the PID algorithms for the decays reconstructed in the LHCb detector.
id cern-2850401
institution Organización Europea para la Investigación Nuclear
language eng
publishDate 2023
record_format invenio
spelling cern-28504012023-06-29T03:38:03Zdoi:10.1088/1742-6596/2438/1/012119http://cds.cern.ch/record/2850401engRyzhikov, ArtemTemirkhanov, AzizDerkach, DenisHushchyn, MikhailKazeev, NikitaMokhnenko, SergeiRobust Neural Particle Identification Modelsphysics.ins-detDetectors and Experimental Techniqueshep-exParticle Physics - ExperimentThe volume of data processed by the Large Hadron Collider experiments demands sophisticated selection rules typically based on machine learning algorithms. One of the shortcomings of these approaches is their profound sensitivity to the biases in training samples. In the case of particle identification (PID), this might lead to degradation of the efficiency for some decays not present in the training dataset due to differences in input kinematic distributions. In this talk, we propose a method based on the Common Specific Decomposition that takes into account individual decays and possible misshapes in the training data by disentangling common and decay specific components of the input feature set. We show that the proposed approach reduces the rate of efficiency degradation for the PID algorithms for the decays reconstructed in the LHCb detector.The volume of data processed by the Large Hadron Collider experiments demands sophisticated selection rules typically based on machine learning algorithms. One of the shortcomings of these approaches is their profound sensitivity to the biases in training samples. In the case of particle identification (PID), this might lead to degradation of the efficiency for some decays not present in the training dataset due to differences in input kinematic distributions. In this talk, we propose a method based on the Common Specific Decomposition that takes into account individual decays and possible misshapes in the training data by disentangling common and decay specific components of the input feature set. We show that the proposed approach reduces the rate of efficiency degradation for the PID algorithms for the decays reconstructed in the LHCb detector.arXiv:2212.07274oai:cds.cern.ch:28504012023
spellingShingle physics.ins-det
Detectors and Experimental Techniques
hep-ex
Particle Physics - Experiment
Ryzhikov, Artem
Temirkhanov, Aziz
Derkach, Denis
Hushchyn, Mikhail
Kazeev, Nikita
Mokhnenko, Sergei
Robust Neural Particle Identification Models
title Robust Neural Particle Identification Models
title_full Robust Neural Particle Identification Models
title_fullStr Robust Neural Particle Identification Models
title_full_unstemmed Robust Neural Particle Identification Models
title_short Robust Neural Particle Identification Models
title_sort robust neural particle identification models
topic physics.ins-det
Detectors and Experimental Techniques
hep-ex
Particle Physics - Experiment
url https://dx.doi.org/10.1088/1742-6596/2438/1/012119
http://cds.cern.ch/record/2850401
work_keys_str_mv AT ryzhikovartem robustneuralparticleidentificationmodels
AT temirkhanovaziz robustneuralparticleidentificationmodels
AT derkachdenis robustneuralparticleidentificationmodels
AT hushchynmikhail robustneuralparticleidentificationmodels
AT kazeevnikita robustneuralparticleidentificationmodels
AT mokhnenkosergei robustneuralparticleidentificationmodels