Interpretable Machine Learning Methods Applied to Jet Background Subtraction in Heavy Ion Collisions

Abstract

Jet measurements in heavy ion collisions can provide constraints on the properties of the quark gluon plasma, but the kinematic reach is limited by a large, fluctuating background. We present a novel application of symbolic regression to extract a functional representation of a deep neural network trained to subtract the background for measurements of jets in relativistic heavy ion collisions. We show that the deep neural network is approximately the same as a method using the particle multiplicity in a jet. This demonstrates that interpretable machine learning methods can provide insight into underlying physical processes.

References in corpus (6)

• Tuning PYTHIA 8.1: the Monash 2013 Tune
• Pileup subtraction using jet areas
• The Catchment Area of Jets
• Measurement of jet fragmentation in PbPb and pp collisions at sqrt(s[NN]) = 2.76 TeV
• Measurement of jet suppression in central Pb-Pb collisions at $\sqrt{s_{\mathrm{NN}}}$ = 2.76 TeV
• Determining the jet transport coefficient $\hat{q}$ from inclusive hadron suppression measurements using Bayesian parameter estimation

Cited by in corpus (6)

• SymbolFit: Automatic Parametric Modeling with Symbolic Regression
• Boosted top tagging and its interpretation using Shapley values
• Accelerating Graph-based Tracking Tasks with Symbolic Regression
• Jet momentum reconstruction in the QGP background with machine learning
• A deep learning approach for predicting multiple observables in Au+Au collisions at RHIC
• Reconstructing jet anisotropies with cumulants