Unsupervised and lightly supervised learning in particle physics

We review the main applications of machine learning models that are not fully supervised in particle physics, i.e., clustering, anomaly detection, detector simulation, and unfolding. Unsupervised methods are ideal for anomaly detection tasks—machine learning models can be trained on background data...

Celý popis

Uloženo v:

Podrobná bibliografie
Vydáno v:	The European physical journal. ST, Special topics Ročník 233; číslo 15-16; s. 2559 - 2596
Hlavní autoři:	Bardhan, Jai, Mandal, Tanumoy, Mitra, Subhadip, Neeraj, Cyrin, Patra, Monalisa
Médium:	Journal Article
Jazyk:	angličtina
Vydáno:	Berlin/Heidelberg Springer Berlin Heidelberg 01.11.2024 Springer Nature B.V
Témata:	Algorithms Anomalies Atomic Classical and Continuum Physics Clustering Condensed Matter Physics Machine learning Materials Science Measurement Science and Instrumentation Modern Machine Learning and Particle Physics: An In-Depth Review Molecular Optical and Plasma Physics Particle physics Partons Physics Physics and Astronomy Review Sensors Simulation Supervised learning Supervision Unsupervised learning Aachen Germany Germany
ISSN:	1951-6355, 1951-6401
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!

Popis
Shrnutí:	We review the main applications of machine learning models that are not fully supervised in particle physics, i.e., clustering, anomaly detection, detector simulation, and unfolding. Unsupervised methods are ideal for anomaly detection tasks—machine learning models can be trained on background data to identify deviations if we model the background data precisely. The learning can also be partially unsupervised when we can provide some information about the anomalies at the data level. Generative models are useful in speeding up detector simulations—they can mimic the computationally intensive task without large resources. They can also efficiently map detector-level data to parton-level data (i.e., data unfolding). In this review, we focus on interesting ideas and connections and briefly overview the underlying techniques wherever necessary.
Bibliografie:	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14
ISSN:	1951-6355 1951-6401
DOI:	10.1140/epjs/s11734-024-01235-x