Orateur
Jérôme Charles
(CPT)
Description
I present a a new model for handling theoretical uncertainties that often complicate the extraction of fundamental parameters from experimental data. The idea is to define a theoretical uncertainty as a fixed but unknow bias, that is let to vary in a given range. The choice of this range as well as its shape in the multidimensional case is discussed with the associated frequentist properties. I illustrate this method with a few examples from indirect tests of the Standard Model.
