A method for tuning parameters of Monte Carlo generators and a its application to the determination of the unintegrated gluon density

TL;DR

The paper introduces a novel parameter tuning method for Monte Carlo generators that uses analytic approximation of observable dependencies to enable faster fitting, demonstrated on unintegrated gluon density fitting with deep inelastic data.

Contribution

It presents a new, efficient parameter tuning technique for Monte Carlo generators using analytic approximations, applied to gluon density determination from experimental data.

Findings

Faster parameter fitting achieved compared to iterative methods.

Successfully fitted unintegrated gluon density using the proposed method.

Discussed the method's limitations and advantages.

Abstract

A method for tuning parameters in Monte Carlo generators is described and applied to a specific case. The method works in the following way: each observable is generated several times using different values of the parameters to be tuned. The output is then approximated by some analytic form to describe the dependence of the observables on the parameters. This approximation is used to find the values of the parameter that give the best description of the experimental data. This results in significantly faster fitting compared to an approach in which the generator is called iteratively. As an application, we employ this method to fit the parameters of the unintegrated gluon density used in the CASCADE Monte Carlo generator, using inclusive deep inelastic data measured by the H1 Collaboration. We discuss the results of the fit, its limitations, and its strong points.