Orateur
Description
Nuclear Parton Distribution Functions (nPDFs) are crucial for understanding nuclear structure and for providing predictions for heavy-ion collisions. nPDFs have been determined via ‘global QCD analyses’, which is a statistical approach based on performing a fit of nPDF-dependent theoretical predictions to the relevant experimental data. One of the crucial aspects of nPDF determination is the estimation of its uncertainties. Typically, the Hessian method is used to propagate experimental uncertainties into predictions for collisions of nuclei. However, due to the nature of nPDF fits (such as limited data constraints, non-gaussianity, and possible multiple minima), this method does not always provide reliable results. Here, we will show a case study for an alternative approach where nPDF uncertainties are estimated using a more advanced statistical method based on the Markov Chain Monte Carlo (MCMC) approach. MCMC methods address nPDF challenges by generating a sequence of random samples from a probability distribution of nPDF parameters and effectively exploring the entire parameter space. This approach allows for a more comprehensive analysis of uncertainties, particularly in complex scenarios such as nPDF fits.
