Studying Hadronization by Machine Learning Techniques

TL;DR

This paper explores the use of deep learning, specifically ResNet neural networks, to model the complex non-perturbative process of hadronization in high-energy physics, aiming to improve predictions of jet and event-shape variables.

Contribution

It introduces a novel application of computer vision and deep learning techniques to study hadronization, providing new insights and tools beyond traditional phenomenological models.

Findings

ResNet networks effectively learn non-linear features of hadronization.

Deep learning models outperform baseline models in predicting jet and event-shape variables.

Results suggest potential for improved modeling of non-perturbative QCD processes.

Abstract

Hadronization is a non-perturbative process, which theoretical description can not be deduced from first principles. Modeling hadron formation requires several assumptions and various phenomenological approaches. Utilizing state-of-the-art Computer Vision and Deep Learning algorithms, it is eventually possible to train neural networks to learn non-linear and non-perturbative features of the physical processes. In this study, results of two ResNet networks are presented by investigating global and kinematical quantities, indeed jet- and event-shape variables. The widely used Lund string fragmentation model is applied as a baseline in $\sqrt{s}= 7$ TeV proton-proton collisions to predict the most relevant observables at further LHC energies.