Playing tag with ANN: Boosted top identification with pattern recognition

Leandro G. Almeida, Mihailo Backović, Mathieu Cliche, Seung Joon Lee, Maxim Perelstein

Research output: Contribution to journalArticle

45 Citations (Scopus)

Abstract

Many searches for physics beyond the Standard Model at the Large Hadron Collider (LHC) rely on top tagging algorithms, which discriminate between boosted hadronic top quarks and the much more common jets initiated by light quarks and gluons. We note that the hadronic calorimeter (HCAL) effectively takes a “digital image” of each jet, with pixel intensities given by energy deposits in individual HCAL cells. Viewed in this way, top tagging becomes a canonical pattern recognition problem. With this motivation, we present a novel top tagging algorithm based on an Artificial Neural Network (ANN), one of the most popular approaches to pattern recognition. The ANN is trained on a large sample of boosted tops and light quark/gluon jets, and is then applied to independent test samples. The ANN tagger demonstrated excellent performance in a Monte Carlo study: for example, for jets with pT in the 1100-1200 GeV range, 60% top-tag efficiency can be achieved with a 4% mis-tag rate. We discuss the physical features of the jets identified by the ANN tagger as the most important for classification, as well as correlations between the ANN tagger and some of the familiar top-tagging observables and algorithms.

Original languageEnglish
Article number86
JournalJournal of High Energy Physics
Volume2015
Issue number7
DOIs
Publication statusPublished - 2015 Jul 27

Fingerprint

pattern recognition
marking
quarks
calorimeters
gluons
pixels
deposits
physics
cells
energy

Keywords

  • Jets

ASJC Scopus subject areas

  • Nuclear and High Energy Physics

Cite this

Playing tag with ANN : Boosted top identification with pattern recognition. / Almeida, Leandro G.; Backović, Mihailo; Cliche, Mathieu; Lee, Seung Joon; Perelstein, Maxim.

In: Journal of High Energy Physics, Vol. 2015, No. 7, 86, 27.07.2015.

Research output: Contribution to journalArticle

Almeida, Leandro G. ; Backović, Mihailo ; Cliche, Mathieu ; Lee, Seung Joon ; Perelstein, Maxim. / Playing tag with ANN : Boosted top identification with pattern recognition. In: Journal of High Energy Physics. 2015 ; Vol. 2015, No. 7.
@article{7fa84b2476e94e689a099cc1a366107a,
title = "Playing tag with ANN: Boosted top identification with pattern recognition",
abstract = "Many searches for physics beyond the Standard Model at the Large Hadron Collider (LHC) rely on top tagging algorithms, which discriminate between boosted hadronic top quarks and the much more common jets initiated by light quarks and gluons. We note that the hadronic calorimeter (HCAL) effectively takes a “digital image” of each jet, with pixel intensities given by energy deposits in individual HCAL cells. Viewed in this way, top tagging becomes a canonical pattern recognition problem. With this motivation, we present a novel top tagging algorithm based on an Artificial Neural Network (ANN), one of the most popular approaches to pattern recognition. The ANN is trained on a large sample of boosted tops and light quark/gluon jets, and is then applied to independent test samples. The ANN tagger demonstrated excellent performance in a Monte Carlo study: for example, for jets with pT in the 1100-1200 GeV range, 60{\%} top-tag efficiency can be achieved with a 4{\%} mis-tag rate. We discuss the physical features of the jets identified by the ANN tagger as the most important for classification, as well as correlations between the ANN tagger and some of the familiar top-tagging observables and algorithms.",
keywords = "Jets",
author = "Almeida, {Leandro G.} and Mihailo Backović and Mathieu Cliche and Lee, {Seung Joon} and Maxim Perelstein",
year = "2015",
month = "7",
day = "27",
doi = "10.1007/JHEP07(2015)086",
language = "English",
volume = "2015",
journal = "Journal of High Energy Physics",
issn = "1126-6708",
publisher = "Springer Verlag",
number = "7",

}

TY - JOUR

T1 - Playing tag with ANN

T2 - Boosted top identification with pattern recognition

AU - Almeida, Leandro G.

AU - Backović, Mihailo

AU - Cliche, Mathieu

AU - Lee, Seung Joon

AU - Perelstein, Maxim

PY - 2015/7/27

Y1 - 2015/7/27

N2 - Many searches for physics beyond the Standard Model at the Large Hadron Collider (LHC) rely on top tagging algorithms, which discriminate between boosted hadronic top quarks and the much more common jets initiated by light quarks and gluons. We note that the hadronic calorimeter (HCAL) effectively takes a “digital image” of each jet, with pixel intensities given by energy deposits in individual HCAL cells. Viewed in this way, top tagging becomes a canonical pattern recognition problem. With this motivation, we present a novel top tagging algorithm based on an Artificial Neural Network (ANN), one of the most popular approaches to pattern recognition. The ANN is trained on a large sample of boosted tops and light quark/gluon jets, and is then applied to independent test samples. The ANN tagger demonstrated excellent performance in a Monte Carlo study: for example, for jets with pT in the 1100-1200 GeV range, 60% top-tag efficiency can be achieved with a 4% mis-tag rate. We discuss the physical features of the jets identified by the ANN tagger as the most important for classification, as well as correlations between the ANN tagger and some of the familiar top-tagging observables and algorithms.

AB - Many searches for physics beyond the Standard Model at the Large Hadron Collider (LHC) rely on top tagging algorithms, which discriminate between boosted hadronic top quarks and the much more common jets initiated by light quarks and gluons. We note that the hadronic calorimeter (HCAL) effectively takes a “digital image” of each jet, with pixel intensities given by energy deposits in individual HCAL cells. Viewed in this way, top tagging becomes a canonical pattern recognition problem. With this motivation, we present a novel top tagging algorithm based on an Artificial Neural Network (ANN), one of the most popular approaches to pattern recognition. The ANN is trained on a large sample of boosted tops and light quark/gluon jets, and is then applied to independent test samples. The ANN tagger demonstrated excellent performance in a Monte Carlo study: for example, for jets with pT in the 1100-1200 GeV range, 60% top-tag efficiency can be achieved with a 4% mis-tag rate. We discuss the physical features of the jets identified by the ANN tagger as the most important for classification, as well as correlations between the ANN tagger and some of the familiar top-tagging observables and algorithms.

KW - Jets

UR - http://www.scopus.com/inward/record.url?scp=84938262433&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84938262433&partnerID=8YFLogxK

U2 - 10.1007/JHEP07(2015)086

DO - 10.1007/JHEP07(2015)086

M3 - Article

AN - SCOPUS:84938262433

VL - 2015

JO - Journal of High Energy Physics

JF - Journal of High Energy Physics

SN - 1126-6708

IS - 7

M1 - 86

ER -