BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:PHYSTAT Seminar: Frequentist Uncertainties on Neural Density Ratio
 s with wifi Ensembles
DTSTART:20260923T140000Z
DTEND:20260923T151500Z
DTSTAMP:20260919T143300Z
UID:indico-event-1702329@indico.cern.ch
DESCRIPTION:Speakers: Sean Benevedes (Georgia Tech)\n\nThis 45'+30' semina
 r is about the paper https://arxiv.org/abs/2506.00113 by Sean Benevedes a
 nd  Jesse Thaler\nAbstract:\nWe introduce wifi ensembles as a novel frame
 work to obtain asymptotic frequentist uncertainties on density ratios\, wi
 th a particular focus on neural ratio estimation in the context of high-en
 ergy physics. When the density ratio of interest is a likelihood ratio con
 ditioned on parameters\, wifi ensembles can be used to perform simulation-
 based inference on those parameters. After training the basis functions f_
 i(x)\, uncertainties on the weights w_i can be straightforwardly propagate
 d to the estimated parameters without requiring extraneous bootstraps. To 
 demonstrate this approach\, we present an application in quantum chromodyn
 amics at the Large Hadron Collider\, using wifi ensembles to estimate the 
 likelihood ratio between generated quark and gluon jets. We use this learn
 ed likelihood ratio to estimate the quark fraction in a synthetic mixed qu
 ark/gluon sample\, showing that the resultant uncertainties empirically sa
 tisfy the desired coverage properties.\n\n\nhttps://indico.cern.ch/event/1
 702329/\n\nZoom: https://cern.zoom.us/j/68793225561?pwd=MHBxOStiUnYvZitTRW
 dvZ05YdkZwUT09
URL:https://indico.cern.ch/event/1702329/
END:VEVENT
END:VCALENDAR