BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Ezequiel Alvarez: Leveraging Biased Simulations with Bayesian infe
 rence
DTSTART:20260602T090000Z
DTEND:20260602T100000Z
DTSTAMP:20260521T111800Z
UID:indico-event-3833@indico.ijs.si
DESCRIPTION:Scientific measurements often yield data that are mixtures of 
 signal and one or more background components. When class templates (e.g.\,
  simulation-based priors) are biased or imprecise\, they are typically tre
 ated as nuisance systematics. We take a different route: by embedding impr
 ecise priors in a hierarchical Bayesian mixture model and combining them w
 ith observed mixed data via Bayesian machine learning\, we adapt class sha
 pes toward the true distributions and obtain an unbiased posterior estimat
 e of the signal fraction with calibrated uncertainties. In practice this l
 ets us exploit simulation even when it is imperfect—improving template f
 idelity during inference. We demonstrate the approach on a collider-physic
 s example (di-Higgs → 4b at the LHC)\,  and show how the same framework
  can enhance and unbias the ABCD background-estimation method.  The frame
 work\, Template-Adapted Mixture Model (TAMM\, arXiv:2604.022219)\, is stra
 ightforward to implement\, scales to high-dimensional and complex distribu
 tions\, and is broadly applicable across HEP\, astrophysics\, cosmology\, 
 and any domain with mixed data and imperfect priors.\n\nhttps://indico.ijs
 .si/event/3833/
LOCATION:F1 tea room
URL:https://indico.ijs.si/event/3833/
END:VEVENT
END:VCALENDAR