Uncertainty quantification for deep learning in particle accelerator applications

With the advent of increased computational resources and improved algorithms, machine learning-based models are being increasingly applied to complex problems in particle accelerators. However, such data-driven models may provide overly confident predictions with unknown errors and uncertainties. Fo...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Aashwin Ananda Mishra, Auralee Edelen, Adi Hanuka, Christopher Mayes
Formato:	article
Lenguaje:	EN
Publicado:	American Physical Society 2021
Materias:	Nuclear and particle physics. Atomic energy. Radioactivity QC770-798
Acceso en línea:	https://doaj.org/article/ad0580f82c8f4e67999e607a714d29f5
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:ad0580f82c8f4e67999e607a714d29f5
record_format	dspace
spelling	oai:doaj.org-article:ad0580f82c8f4e67999e607a714d29f52021-11-29T18:17:49ZUncertainty quantification for deep learning in particle accelerator applications10.1103/PhysRevAccelBeams.24.1146012469-9888https://doaj.org/article/ad0580f82c8f4e67999e607a714d29f52021-11-01T00:00:00Zhttp://doi.org/10.1103/PhysRevAccelBeams.24.114601http://doi.org/10.1103/PhysRevAccelBeams.24.114601https://doaj.org/toc/2469-9888With the advent of increased computational resources and improved algorithms, machine learning-based models are being increasingly applied to complex problems in particle accelerators. However, such data-driven models may provide overly confident predictions with unknown errors and uncertainties. For reliable deployment of machine learning models in high-regret and safety-critical systems such as particle accelerators, estimates of prediction uncertainty are needed along with accurate point predictions. In this investigation, we evaluate Bayesian neural networks (BNN) as an approach that can provide accurate predictions along with reliably quantified uncertainties for particle accelerator problems, and compare their performance with bootstrapped ensembles of neural networks. We select three accelerator setups for this evaluation: a storage ring, a photoinjector, and a linac. The problems span different data volumes and dimensionalities (e.g., scalar predictions as well as image outputs). It is found that BNN provide accurate predictions of the mean along with reliable estimates of predictive uncertainty across the test cases. In this vein, BNN may offer an attractive alternative to deterministic deep learning tools to generate accurate predictions with quantified uncertainties in particle accelerator applications.Aashwin Ananda MishraAuralee EdelenAdi HanukaChristopher MayesAmerican Physical SocietyarticleNuclear and particle physics. Atomic energy. RadioactivityQC770-798ENPhysical Review Accelerators and Beams, Vol 24, Iss 11, p 114601 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Nuclear and particle physics. Atomic energy. Radioactivity QC770-798
spellingShingle	Nuclear and particle physics. Atomic energy. Radioactivity QC770-798 Aashwin Ananda Mishra Auralee Edelen Adi Hanuka Christopher Mayes Uncertainty quantification for deep learning in particle accelerator applications
description	With the advent of increased computational resources and improved algorithms, machine learning-based models are being increasingly applied to complex problems in particle accelerators. However, such data-driven models may provide overly confident predictions with unknown errors and uncertainties. For reliable deployment of machine learning models in high-regret and safety-critical systems such as particle accelerators, estimates of prediction uncertainty are needed along with accurate point predictions. In this investigation, we evaluate Bayesian neural networks (BNN) as an approach that can provide accurate predictions along with reliably quantified uncertainties for particle accelerator problems, and compare their performance with bootstrapped ensembles of neural networks. We select three accelerator setups for this evaluation: a storage ring, a photoinjector, and a linac. The problems span different data volumes and dimensionalities (e.g., scalar predictions as well as image outputs). It is found that BNN provide accurate predictions of the mean along with reliable estimates of predictive uncertainty across the test cases. In this vein, BNN may offer an attractive alternative to deterministic deep learning tools to generate accurate predictions with quantified uncertainties in particle accelerator applications.
format	article
author	Aashwin Ananda Mishra Auralee Edelen Adi Hanuka Christopher Mayes
author_facet	Aashwin Ananda Mishra Auralee Edelen Adi Hanuka Christopher Mayes
author_sort	Aashwin Ananda Mishra
title	Uncertainty quantification for deep learning in particle accelerator applications
title_short	Uncertainty quantification for deep learning in particle accelerator applications
title_full	Uncertainty quantification for deep learning in particle accelerator applications
title_fullStr	Uncertainty quantification for deep learning in particle accelerator applications
title_full_unstemmed	Uncertainty quantification for deep learning in particle accelerator applications
title_sort	uncertainty quantification for deep learning in particle accelerator applications
publisher	American Physical Society
publishDate	2021
url	https://doaj.org/article/ad0580f82c8f4e67999e607a714d29f5
work_keys_str_mv	AT aashwinanandamishra uncertaintyquantificationfordeeplearninginparticleacceleratorapplications AT auraleeedelen uncertaintyquantificationfordeeplearninginparticleacceleratorapplications AT adihanuka uncertaintyquantificationfordeeplearninginparticleacceleratorapplications AT christophermayes uncertaintyquantificationfordeeplearninginparticleacceleratorapplications
_version_	1718407182715715584

Uncertainty quantification for deep learning in particle accelerator applications

Ejemplares similares