Cargando…

Assessment of few-hits machine learning classification algorithms for low energy physics in liquid argon detectors

The physics potential of massive liquid argon TPCs in the low-energy regime is still to be fully reaped because few-hits events encode information that can hardly be exploited by conventional classification algorithms. Machine learning (ML) techniques give their best in these types of classification...

Descripción completa

Detalles Bibliográficos
Autores principales: Biassoni, Matteo, Giachero, Andrea, Grossi, Michele, Guffanti, Daniele, Labranca, Danilo, Moretti, Roberto, Rossi, Marco, Terranova, Francesco, Vallecorsa, Sofia
Lenguaje:eng
Publicado: 2023
Materias:
Acceso en línea:http://cds.cern.ch/record/2861063
_version_ 1780977791500550144
author Biassoni, Matteo
Giachero, Andrea
Grossi, Michele
Guffanti, Daniele
Labranca, Danilo
Moretti, Roberto
Rossi, Marco
Terranova, Francesco
Vallecorsa, Sofia
author_facet Biassoni, Matteo
Giachero, Andrea
Grossi, Michele
Guffanti, Daniele
Labranca, Danilo
Moretti, Roberto
Rossi, Marco
Terranova, Francesco
Vallecorsa, Sofia
author_sort Biassoni, Matteo
collection CERN
description The physics potential of massive liquid argon TPCs in the low-energy regime is still to be fully reaped because few-hits events encode information that can hardly be exploited by conventional classification algorithms. Machine learning (ML) techniques give their best in these types of classification problems. In this paper, we evaluate their performance against conventional (deterministic) algorithms. We demonstrate that both Convolutional Neural Networks (CNN) and Transformer-Encoder methods outperform deterministic algorithms in one of the most challenging classification problems of low-energy physics (single- versus double-beta events). We discuss the advantages and pitfalls of Transformer-Encoder methods versus CNN and employ these methods to optimize the detector parameters, with an emphasis on the DUNE Phase II detectors ("Module of Opportunity").
id cern-2861063
institution Organización Europea para la Investigación Nuclear
language eng
publishDate 2023
record_format invenio
spelling cern-28610632023-09-27T07:58:53Zhttp://cds.cern.ch/record/2861063engBiassoni, MatteoGiachero, AndreaGrossi, MicheleGuffanti, DanieleLabranca, DaniloMoretti, RobertoRossi, MarcoTerranova, FrancescoVallecorsa, SofiaAssessment of few-hits machine learning classification algorithms for low energy physics in liquid argon detectorsphysics.data-anOther Fields of Physicscs.LGComputing and Computersphysics.ins-detDetectors and Experimental TechniquesThe physics potential of massive liquid argon TPCs in the low-energy regime is still to be fully reaped because few-hits events encode information that can hardly be exploited by conventional classification algorithms. Machine learning (ML) techniques give their best in these types of classification problems. In this paper, we evaluate their performance against conventional (deterministic) algorithms. We demonstrate that both Convolutional Neural Networks (CNN) and Transformer-Encoder methods outperform deterministic algorithms in one of the most challenging classification problems of low-energy physics (single- versus double-beta events). We discuss the advantages and pitfalls of Transformer-Encoder methods versus CNN and employ these methods to optimize the detector parameters, with an emphasis on the DUNE Phase II detectors ("Module of Opportunity").arXiv:2305.09744oai:cds.cern.ch:28610632023-05-16
spellingShingle physics.data-an
Other Fields of Physics
cs.LG
Computing and Computers
physics.ins-det
Detectors and Experimental Techniques
Biassoni, Matteo
Giachero, Andrea
Grossi, Michele
Guffanti, Daniele
Labranca, Danilo
Moretti, Roberto
Rossi, Marco
Terranova, Francesco
Vallecorsa, Sofia
Assessment of few-hits machine learning classification algorithms for low energy physics in liquid argon detectors
title Assessment of few-hits machine learning classification algorithms for low energy physics in liquid argon detectors
title_full Assessment of few-hits machine learning classification algorithms for low energy physics in liquid argon detectors
title_fullStr Assessment of few-hits machine learning classification algorithms for low energy physics in liquid argon detectors
title_full_unstemmed Assessment of few-hits machine learning classification algorithms for low energy physics in liquid argon detectors
title_short Assessment of few-hits machine learning classification algorithms for low energy physics in liquid argon detectors
title_sort assessment of few-hits machine learning classification algorithms for low energy physics in liquid argon detectors
topic physics.data-an
Other Fields of Physics
cs.LG
Computing and Computers
physics.ins-det
Detectors and Experimental Techniques
url http://cds.cern.ch/record/2861063
work_keys_str_mv AT biassonimatteo assessmentoffewhitsmachinelearningclassificationalgorithmsforlowenergyphysicsinliquidargondetectors
AT giacheroandrea assessmentoffewhitsmachinelearningclassificationalgorithmsforlowenergyphysicsinliquidargondetectors
AT grossimichele assessmentoffewhitsmachinelearningclassificationalgorithmsforlowenergyphysicsinliquidargondetectors
AT guffantidaniele assessmentoffewhitsmachinelearningclassificationalgorithmsforlowenergyphysicsinliquidargondetectors
AT labrancadanilo assessmentoffewhitsmachinelearningclassificationalgorithmsforlowenergyphysicsinliquidargondetectors
AT morettiroberto assessmentoffewhitsmachinelearningclassificationalgorithmsforlowenergyphysicsinliquidargondetectors
AT rossimarco assessmentoffewhitsmachinelearningclassificationalgorithmsforlowenergyphysicsinliquidargondetectors
AT terranovafrancesco assessmentoffewhitsmachinelearningclassificationalgorithmsforlowenergyphysicsinliquidargondetectors
AT vallecorsasofia assessmentoffewhitsmachinelearningclassificationalgorithmsforlowenergyphysicsinliquidargondetectors