BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Hessian PDF reweighting meets the Bayesian methods
DTSTART;VALUE=DATE-TIME:20140430T143000Z
DTEND;VALUE=DATE-TIME:20140430T145500Z
DTSTAMP;VALUE=DATE-TIME:20191019T233738Z
UID:indico-contribution-1588306@indico.cern.ch
DESCRIPTION:Speakers: Hannu Paukkunen (University of Jyväskylä)\nWe disc
uss the Hessian PDF reweighting - a technique intended to estimate the eff
ects that new measurements have on a set of PDFs. The method stems straig
htforwardly from considering new data in a usual $\\chi^2$ fit and it natu
rally incorporates also non-zero values for the tolerance\, $\\Delta \\chi
^2 > 1$. In comparison to the contemporary Bayesian reweighting techniques
\, there is no need to generate large ensembles of PDF Monte-Carlo replica
s\, and the observables need to be evaluated only with the central and the
error sets of the original PDFs. In spite of the apparently rather differ
ent methodologies\, we show that the Hessian and the Bayesian techniques a
re actually one and the same\, but only if the $\\Delta \\chi^2$ criterion
is properly included to the Bayesian likelihood function that is a simple
exponential. We illustrate the situation by considering a simplified exam
ple and the case of inclusive jets at the LHC.\n\nhttps://indico.cern.ch/e
vent/258017/contributions/1588306/
LOCATION:BUW Auditorium
URL:https://indico.cern.ch/event/258017/contributions/1588306/
END:VEVENT
END:VCALENDAR