Deep Learning Jet Substructure from Two-Particle Correlation

TL;DR

This paper introduces a novel two-particle correlation neural network (2PCNN) for analyzing jet substructure in collider physics, integrating particle data to improve tagging performance and interpret physical features.

Contribution

It presents a new neural network architecture that combines two-particle correlations with deep learning for jet analysis, enhancing interpretability and performance.

Findings

Achieves superior boosted boson and heavy flavor tagging performance.

Identifies key correlation pairs with physical significance.

Provides insights into jet substructure through learned correlations.

Abstract

Deciphering the complex information contained in jets produced in collider events requires a physical organization of the jet data. We introduce two-particle correlations (2PCs) by pairing individual particles as the initial jet representation from which a probabilistic model can be built. Particle momenta, as well as particle types and vertex information are included in the correlation. A novel, two-particle correlation neural network (2PCNN) architecture is constructed by combining neural network based filters on 2PCs and a deep neural network for capturing jet kinematic information. The 2PCNN is applied to boosted boson and heavy flavor tagging, and it achieves excellent performance by comparing to models based on telescoping deconstruction. Major correlation pairs exploited in the trained models are also identified, which shed light on the physical significance of certain jet substructure.