Deep learning to improve the sensitivity of Di-Higgs searches in the 4b channel

Cheng Wei Chiang; Feng Yang Hsieh; Shih Chieh Hsu; Ian Low

doi:10.1007/JHEP09(2024)139

Deep learning to improve the sensitivity of Di-Higgs searches in the 4b channel

Cheng Wei Chiang, Feng Yang Hsieh^*, Shih Chieh Hsu, Ian Low

^*Corresponding author for this work

Physics and Astronomy

Research output: Contribution to journal › Article › peer-review

Abstract

The study of di-Higgs events, both resonant and non-resonant, plays a crucial role in understanding the fundamental interactions of the Higgs boson. In this work we consider di-Higgs events decaying into four b-quarks and propose to improve the experimental sensitivity by utilizing a novel machine learning algorithm known as Symmetry Preserving Attention Network (Spa-Net) — a neural network structure whose architecture is designed to incorporate the inherent symmetries in particle reconstruction tasks. We demonstrate that the Spa-Net can enhance the experimental reach over baseline methods such as the cut-based and the Dense Neural Network-based analyses. At the Large Hadron Collider, with a 14-TeV center-of-mass energy and an integrated luminosity of 300 fb⁻¹, the Spa-Net allows us to establish 95% C.L. upper limits in resonant production cross-sections that are 10% to 45% stronger than baseline methods. For non-resonant di-Higgs production, Spa-Net enables us to constrain the self-coupling that is 9% more stringent than the baseline method.

Original language	English (US)
Article number	139
Journal	Journal of High Energy Physics
Volume	2024
Issue number	9
DOIs	https://doi.org/10.1007/JHEP09(2024)139
State	Published - Sep 2024

Funding

We are grateful to Alexander Shmakov for the assistance with the Spa-Net package. Additionally, we extend special thanks to David Shih and Alexander Shmakov for their valuable comments on our manuscript. C.-W. Chiang and F.-Y. Hsieh are supported in part by the National Science and Technology Council of Taiwan under Grant No. NSTC-111-2112-M-002-018-MY3. S.-C. Hsu is supported by the U.S. National Science Foundation grants No. 2110963. Work at Argonne is supported in part by the U.S. Department of Energy under contract DE-AC02-06CH11357. I. Low acknowledges the hospitality of the Phenomenology Group at National Taiwan University during the completion of this work.

Keywords

Higgs Production
Higgs Properties

ASJC Scopus subject areas

Nuclear and High Energy Physics

Access to Document

10.1007/JHEP09(2024)139

Cite this

@article{4e6729bf1a8443f1961bb0403d489f87,

title = "Deep learning to improve the sensitivity of Di-Higgs searches in the 4b channel",

abstract = "The study of di-Higgs events, both resonant and non-resonant, plays a crucial role in understanding the fundamental interactions of the Higgs boson. In this work we consider di-Higgs events decaying into four b-quarks and propose to improve the experimental sensitivity by utilizing a novel machine learning algorithm known as Symmetry Preserving Attention Network (Spa-Net) — a neural network structure whose architecture is designed to incorporate the inherent symmetries in particle reconstruction tasks. We demonstrate that the Spa-Net can enhance the experimental reach over baseline methods such as the cut-based and the Dense Neural Network-based analyses. At the Large Hadron Collider, with a 14-TeV center-of-mass energy and an integrated luminosity of 300 fb−1, the Spa-Net allows us to establish 95% C.L. upper limits in resonant production cross-sections that are 10% to 45% stronger than baseline methods. For non-resonant di-Higgs production, Spa-Net enables us to constrain the self-coupling that is 9% more stringent than the baseline method.",

keywords = "Higgs Production, Higgs Properties",

author = "Chiang, {Cheng Wei} and Hsieh, {Feng Yang} and Hsu, {Shih Chieh} and Ian Low",

note = "Publisher Copyright: {\textcopyright} The Author(s) 2024.",

year = "2024",

month = sep,

doi = "10.1007/JHEP09(2024)139",

language = "English (US)",

volume = "2024",

journal = "Journal of High Energy Physics",

issn = "1126-6708",

publisher = "Springer Verlag",

number = "9",

}

TY - JOUR

T1 - Deep learning to improve the sensitivity of Di-Higgs searches in the 4b channel

AU - Chiang, Cheng Wei

AU - Hsieh, Feng Yang

AU - Hsu, Shih Chieh

AU - Low, Ian

PY - 2024/9

Y1 - 2024/9

N2 - The study of di-Higgs events, both resonant and non-resonant, plays a crucial role in understanding the fundamental interactions of the Higgs boson. In this work we consider di-Higgs events decaying into four b-quarks and propose to improve the experimental sensitivity by utilizing a novel machine learning algorithm known as Symmetry Preserving Attention Network (Spa-Net) — a neural network structure whose architecture is designed to incorporate the inherent symmetries in particle reconstruction tasks. We demonstrate that the Spa-Net can enhance the experimental reach over baseline methods such as the cut-based and the Dense Neural Network-based analyses. At the Large Hadron Collider, with a 14-TeV center-of-mass energy and an integrated luminosity of 300 fb−1, the Spa-Net allows us to establish 95% C.L. upper limits in resonant production cross-sections that are 10% to 45% stronger than baseline methods. For non-resonant di-Higgs production, Spa-Net enables us to constrain the self-coupling that is 9% more stringent than the baseline method.

AB - The study of di-Higgs events, both resonant and non-resonant, plays a crucial role in understanding the fundamental interactions of the Higgs boson. In this work we consider di-Higgs events decaying into four b-quarks and propose to improve the experimental sensitivity by utilizing a novel machine learning algorithm known as Symmetry Preserving Attention Network (Spa-Net) — a neural network structure whose architecture is designed to incorporate the inherent symmetries in particle reconstruction tasks. We demonstrate that the Spa-Net can enhance the experimental reach over baseline methods such as the cut-based and the Dense Neural Network-based analyses. At the Large Hadron Collider, with a 14-TeV center-of-mass energy and an integrated luminosity of 300 fb−1, the Spa-Net allows us to establish 95% C.L. upper limits in resonant production cross-sections that are 10% to 45% stronger than baseline methods. For non-resonant di-Higgs production, Spa-Net enables us to constrain the self-coupling that is 9% more stringent than the baseline method.

KW - Higgs Production

KW - Higgs Properties

UR - http://www.scopus.com/inward/record.url?scp=85204483291&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85204483291&partnerID=8YFLogxK

U2 - 10.1007/JHEP09(2024)139

DO - 10.1007/JHEP09(2024)139

M3 - Article

AN - SCOPUS:85204483291

SN - 1126-6708

VL - 2024

JO - Journal of High Energy Physics

JF - Journal of High Energy Physics

IS - 9

M1 - 139

ER -

Deep learning to improve the sensitivity of Di-Higgs searches in the 4b channel

Abstract

Funding

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this