BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Cavendish HEP Seminars: Errors-on-Errors and Higher-Order Asymptot
 ics in Particle Physics
DTSTART:20261027T110000Z
DTEND:20261027T121500Z
DTSTAMP:20261024T005400Z
UID:indico-event-19565@indico.global
DESCRIPTION:Speakers: Nikolai Fomin (University of Cambridge (GB))\, Matth
 ew David Monk (University of Cambridge (GB))\n\nhttps://talks.cam.ac.uk/ta
 lk/index/273082/\n \nIn particle physics\, data are collected and compare
 d to theoretical predictions by constructing a statistical model that capt
 ures both the intrinsic stochastic nature of the measurement process and p
 otential biases introduced by approximations in theoretical predictions an
 d in the modeling of the experimental setup\, referred to as systematic un
 certainties. However\, these uncertainties are often derived from limited 
 data or estimated through ad-hoc procedures\, introducing uncertainty in t
 he uncertainty itself - informally referred to as ``error-on-error’'.\n
  \nIn this talk a statistical framework to address this issue is presente
 d. This is achieved by explicitly treating systematic uncertainties as imp
 erfect estimates of unknown error parameters\, which effectively incorpora
 tes errors-on-errors into the statistical model. This formulation naturall
 y reduces sensitivity to outliers and captures internal tensions within da
 tasets by interpreting them as an additional source of uncertainty. After 
 a formal introduction to the model\, we illustrate its application to seve
 ral particle-physics problems\, including the determination of parton dist
 ribution functions.\n \nThe inclusion of errors-on-errors can also limit 
 the accuracy of standard asymptotic approximations\, motivating higher-ord
 er corrections for goodness-of-fit tests and confidence intervals. This co
 nnects naturally to discovery searches with small event yields\, where lar
 ge-sample approximations can likewise become inaccurate. Counting experime
 nts with backgrounds estimated from control measurements provide a simple 
 setting in which to examine these limitations and demonstrate how higher-o
 rder methods improve estimates of observed and expected discovery signific
 ance.\n\nhttps://indico.global/event/19565/
LOCATION:Ray Dolby Centre Seminar Room B\, B2.002
URL:https://indico.global/event/19565/
END:VEVENT
END:VCALENDAR