Inference from Relativistic Heavy-Ion Data

Colliding large nuclei at velocities approaching the speed of light produces a plasma of strongly interacting nuclear matter known as quark-gluon plasma. This nuclear plasma can be characterized by macroscopic properties such as its equation of state and viscosity. Bayesian inference provides a systematic framework for constraining these properties using experimental measurements from the Relativistic Heavy Ion Collider and the Large Hadron Collider. This presentation surveys diverse applications of Bayesian methodology in heavy-ion collision analysis, including transfer learning, model averaging, stochastic emulator uncertainty optimization, and construction of covariance error matrices from limited information. Multiple results from the JETSCAPE Collaboration will be discussed.