Quark-Gluon Jet Discrimination Using Convolutional Neural Networks

TL;DR

This paper explores the use of convolutional neural networks to differentiate quark-initiated jets from gluon-initiated jets in collider data by treating jets as images, comparing CNN performance with traditional methods.

Contribution

It demonstrates the effectiveness of state-of-the-art CNN architectures for quark-gluon jet discrimination, advancing particle physics data analysis techniques.

Findings

CNNs outperform traditional BDT in jet classification accuracy

Different CNN architectures show varying levels of discrimination performance

Treating jets as images is a promising approach for collider data analysis

Abstract

Currently, newly developed artificial intelligence techniques, in particular convolutional neural networks, are being investigated for use in data-processing and classification of particle physics collider data. One such challenging task is to distinguish quark-initiated jets from gluon-initiated jets. Following previous work, we treat the jet as an image by pixelizing track information and calorimeter deposits as reconstructed by the detector. We test the deep learning paradigm by training several recently developed, state-of-the-art convolutional neural networks on the quark-gluon discrimination task. We compare the results obtained using various network architectures trained for quark-gluon discrimination and also a boosted decision tree (BDT) trained on summary variables.