Jet Constituents for Deep Neural Network Based Top Quark Tagging

TL;DR

This paper introduces a sequential jet constituent approach for deep neural network top quark tagging, preserving detailed information and achieving high background rejection across a broad momentum range.

Contribution

It presents a novel sequential input method for jet tagging that maintains full information and improves performance over traditional image or high-level feature techniques.

Findings

Achieves background rejection of 45 at 50% efficiency

Insensitive to multiple proton-proton interactions

Effective for jets with transverse momentum 600-2500 GeV

Abstract

Recent literature on deep neural networks for tagging of highly energetic jets resulting from top quark decays has focused on image based techniques or multivariate approaches using high-level jet substructure variables. Here, a sequential approach to this task is taken by using an ordered sequence of jet constituents as training inputs. Unlike the majority of previous approaches, this strategy does not result in a loss of information during pixelisation or the calculation of high level features. The jet classification method achieves a background rejection of 45 at a 50% efficiency operating point for reconstruction level jets with transverse momentum range of 600 to 2500 GeV and is insensitive to multiple proton-proton interactions at the levels expected throughout Run 2 of the LHC.