CaltechAUTHORS
Published July 1, 2020 | Version Submitted + Published
Journal Article Open

Interaction networks for the identification of boosted H→bb̅ decays

Creators

  • 1. ROR icon California Institute of Technology
  • 2. ROR icon University of California, San Diego
  • 3. ROR icon European Organization for Nuclear Research

Abstract

We develop a jet identification algorithm based on an interaction network, designed to identify high-momentum Higgs bosons decaying to bottom quark-antiquark pairs, distinguish them from ordinary jets originating from the hadronization of quarks and gluons. The algorithm's inputs are features of the reconstructed charged particles in a jet and the secondary vertices associated to them. Describing the jet shower as a combination of particle-to-particle and particle-to-vertex interactions, the model is trained to learn a jet representation on which the classification problem is optimized. The algorithm is trained on simulated samples of accurate LHC collisions, released by the CMS collaboration on the CERN Open Data Portal. The interaction network achieves a drastic improvement in the identification performance with respect to state-of-the-art algorithms.

Additional Information

© 2020 Published by the American Physical Society under the terms of the Creative Commons Attribution 4.0 International license. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI. Funded by SCOAP3. Received 10 December 2019; accepted 13 June 2020; published 28 July 2020. This work was possible thanks to the commitment of the CMS Collaboration to release its simulation data through the CERN Open Data Portal. We would like to thank our CMS colleagues and the CERN Open Data team for their effort to promote open access to science. In particular, we thank Kati Lassila-Perini for her precious help. We are grateful to Caltech and the Kavli Foundation for their support of undergraduate student research in cross-cutting areas of machine learning and domain sciences. This work was conducted at "iBanks," the AI GPU cluster at Caltech. We acknowledge NVIDIA, SuperMicro and the Kavli Foundation for their support of iBanks. This project is partially supported by the United States Department of Energy, Office of High Energy Physics Research under Caltech Contract No. DE-SC0011925. M. P. is supported by the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program (Grant Agreement No. 772369). J. M. D. is supported by Fermi Research Alliance, LLC under Contract No. DE-AC02-07CH11359 with the U.S. Department of Energy, Office of Science, Office of High Energy Physics. E. A. M is supported by the Taylor W. Lawrence Research Fellowship and Mellon Mays Research Fellowship.

Attached Files

Published - PhysRevD.102.012010.pdf

Submitted - 1909.12285.pdf

Files

1909.12285.pdf

Files (3.2 MB)

Name Size
md5:e7374083fd8b34197b899c7fcd795c0d
1.6 MB Preview Download
md5:b9bd0029ebc4f10775da2da9e421a511
1.6 MB Preview Download

Additional details

Identifiers

Eprint ID
99386
Resolver ID
CaltechAUTHORS:20191022-101506524

Related works

Describes
https://arxiv.org/abs/1909.12285 (URL)

Funding

Caltech
Kavli Foundation
NVIDIA
SuperMicro
Department of Energy (DOE)
DE-SC0011925
European Research Council (ERC)
772369
Department of Energy (DOE)
DE-AC02-07CH11359
Taylor W. Lawrence Research Fellowship
Mellon Mays Research Fellowship
SCOAP3

Dates

Created
2019-10-22
Created from EPrint's datestamp field
Updated
2021-11-16
Created from EPrint's last_modified field

Caltech Custom Metadata

Caltech groups
CMS@Caltech , Physics Department