Bayesian quantification of the Quark-Gluon Plasma: Improved design and closure demonstration

TL;DR

This paper demonstrates a comprehensive Bayesian framework for comparing heavy-ion collision models with experimental data, introducing improved sampling methods to better explore model parameters and validate predictions.

Contribution

It introduces a systematically-improvable sampling design for Bayesian model-to-data comparison in heavy-ion physics, enhancing parameter exploration and model validation.

Findings

First comprehensive Bayesian comparison of heavy-ion data with IP-Glasma and hydrodynamics.

Development of improved sampling design with better projection properties.

Demonstration of model closure and validation techniques.

Abstract

We present a demonstration of the design sampling and closure for the first comprehensive Bayesian model-to-data comparison of heavy-ion measurements with IP-Glasma initial conditions, in which we combine with state-of-the-art hydrodynamics (MUSIC), particlization (iS3D), and transport (SMASH). We further introduce a systematically-improvable method of sampling design points with better projection properties to explore the parameter space of the model for the first time in heavy-ion collisions.