BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:[RBRC seminar] Inferring Parton Distributions: Bayesian & Deep Gen
 erative Modeling
DTSTART:20260618T163000Z
DTEND:20260618T183000Z
DTSTAMP:20260604T080900Z
UID:indico-event-32625@indico.bnl.gov
DESCRIPTION:Speakers: Yamil Cahuana Medrano (College of William & Mary)\n\
 nExtracting parton distribution functions (PDFs) from Lattice QCD is cruci
 al for understanding nucleon structure\, but it fundamentally relies on so
 lving a challenging ill-posed inverse problem. In this talk\, I will prese
 nt an overview of my work on solving this problem within the pseudo-PDF fr
 amework. First\, I will outline how Gaussian processes (GPs) provide flexi
 ble Bayesian priors that encode correlations and physical constraints with
 out fixing a functional shape\, quantifying information gain through the K
 ullback–Leibler divergence. Additionally\, I will present an alternative
  non-parametric strategy based on deep generative modeling\, utilizing inv
 ertible neural networks (INNs) conditioned on a GP prior. Tests on synthet
 ic data confirm the consistency and robustness of both independent methodo
 logies. These results support both GP regression and INN frameworks as sys
 tematic approaches to PDF reconstruction\, offering controlled uncertainti
 es and reduced model bias in Lattice QCD analyses.\n\nhttps://indico.bnl.g
 ov/event/32625/
LOCATION:2-160 (https://bnl.zoomgov.com/j/1600983728?pwd=RAD7OLcqre7Ycsp6J
 fFp6HAnpyLxex.1)
URL:https://indico.bnl.gov/event/32625/
END:VEVENT
END:VCALENDAR