Salvato in:
Dettagli Bibliografici
Autori principali: Vigl, Matthias, Hartman, Nicole, Heinrich, Lukas
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2401.13536
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909082669023232
author Vigl, Matthias
Hartman, Nicole
Heinrich, Lukas
author_facet Vigl, Matthias
Hartman, Nicole
Heinrich, Lukas
contents In this work we demonstrate that significant gains in performance and data efficiency can be achieved in High Energy Physics (HEP) by moving beyond the standard paradigm of sequential optimization or reconstruction and analysis components. We conceptually connect HEP reconstruction and analysis to modern machine learning workflows such as pretraining, finetuning, domain adaptation and high-dimensional embedding spaces and quantify the gains in the example usecase of searches of heavy resonances decaying via an intermediate di-Higgs system to four $b$-jets.
format Preprint
id arxiv_https___arxiv_org_abs_2401_13536
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Finetuning Foundation Models for Joint Analysis Optimization
Vigl, Matthias
Hartman, Nicole
Heinrich, Lukas
High Energy Physics - Experiment
Machine Learning
High Energy Physics - Phenomenology
Data Analysis, Statistics and Probability
In this work we demonstrate that significant gains in performance and data efficiency can be achieved in High Energy Physics (HEP) by moving beyond the standard paradigm of sequential optimization or reconstruction and analysis components. We conceptually connect HEP reconstruction and analysis to modern machine learning workflows such as pretraining, finetuning, domain adaptation and high-dimensional embedding spaces and quantify the gains in the example usecase of searches of heavy resonances decaying via an intermediate di-Higgs system to four $b$-jets.
title Finetuning Foundation Models for Joint Analysis Optimization
topic High Energy Physics - Experiment
Machine Learning
High Energy Physics - Phenomenology
Data Analysis, Statistics and Probability
url https://arxiv.org/abs/2401.13536