Bayesian Optimization of Pythia8 Tunes

TL;DR

This paper applies Bayesian optimization to tune Pythia8's parameters, achieving a better fit to LEPI data than default settings, marking a comprehensive use of this method in parton shower and hadronization modeling.

Contribution

It introduces the first comprehensive Bayesian optimization approach for tuning Pythia8's key parameters using LEPI data.

Findings

The new tune outperforms the default in fitting LEPI data.

Bayesian optimization effectively tunes complex particle physics models.

This approach enhances the precision of Pythia8 simulations.

Abstract

A new tune (set of model parameters) is found for the six most important parameters of the Pythia8 final state parton shower and hadronization model using Bayesian optimization. The tune fits the LEPI data from ALEPH better than the default tune in Pythia8. To the best of our knowledge, we present the most comprehensive application of Bayesian optimization to the tuning of a parton shower and hadronization model using the LEPI data.