Heidelberg 2022 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
T: Fachverband Teilchenphysik
T 58: QCD (Exp.) 2
T 58.3: Vortrag
Mittwoch, 23. März 2022, 16:45–17:00, T-H17
Studies on Monte Carlo tuning using Bayesian Analysis — •Salvatore La Cagnina1, Andrii Verbytski2, Johannes Erdmann1, Kevin Kröninger1, and Stefan Kluth2 — 1TU Dortmund, Fakultät Physik — 2Max-Plank-Institut für Physik, München
Monte Carlo (MC) simulations are an essential aspect of data analysis at the LHC. One aspect of MC event generation involves hadronisation and parton shower models. Since these models are based on approximations, they introduce a number of parameters. These parameters cannot be inferred from first principles. Therefore, their values have to be optimized using numerical tools and experimental data (MC tuning). Generally, MC tuning is performed by choosing observables that are sensitive to the parameters. Afterwards, a fit of the parameters to data using a simplified MC response function derived from fits to MC events is performed. Though state-of-the-art methods for MC tuning exist, uncertainties are usually treated as uncorrelated. In this talk, MC tuning using a Bayesian approach will be discussed. The EFTfitter tool is used for fitting, which enables the implementation of correlations for different sources of uncertainties. First results using this method on a MC tune with LEP data will be presented.