BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20250713T085909EDT-40584fS50d@132.216.98.100
DTSTAMP:20250713T125909Z
DESCRIPTION:Title: Distribution-​free inference for regression: discrete\, 
 continuous\, and in between.\n\n\n	Abstract:\n\n\nIn data analysis problems
  where we are not able to rely on distributional assumptions\, what types 
 of inference guarantees can still be obtained? Many popular methods\, such
  as holdout methods\, cross-validation methods\, and conformal prediction\
 , are able to provide distribution-free guarantees for predictive inferenc
 e\, but the problem of providing inference for the underlying regression f
 unction (for example\, inference on the conditional mean E[Y|X]) is more c
 hallenging. If X takes only a small number of possible values\, then infer
 ence on E[Y|X] is trivial to achieve. At the other extreme\, if the featur
 es X are continuously distributed\, we show that any confidence interval f
 or E[Y|X] must have non-vanishing width\, even as sample size tends to inf
 inity - this is true regardless of smoothness properties or other desirabl
 e features of the underlying distribution. In between these two extremes\,
  we find several distinct regimes - in particular\, it is possible for dis
 tribution-free confidence intervals to have vanishing width if and only if
  the effective support size of the distribution ofXis smaller than the squ
 are of the sample size.\n\nThis work is joint with Yonghoon Lee.\n\n\n	Spea
 ker\n\n\nDr. Barber is the Louis Block Professor of statistics at the Univ
 ersity of Chicago. Before starting at U of C\, she worked in the Departmen
 t of Statistics at Stanford University\, supervised by Dr. Emmanuel Candes
 . She was awarded the 2020 COPSS Presidents' Award for her fundamental con
 tributions to statistical sparsity and selective inference in high-dimensi
 onal problems\; for the creative and novel knockoff filter to cope with co
 rrelated coefficients\; for contributions to compressed sensing\, the jack
 knife\, and conformal predictive inference\; and for the encouragement and
  training of graduate and undergraduate students.\n\nHer research interest
 s are in developing and analyzing estimation\, inference\, and optimizatio
 n tools for structured high-dimensional data problems such as sparse regre
 ssion\, sparse nonparametric models\, and low-rank models. She also collab
 orate on modeling and optimization problems in image reconstruction for me
 dical imaging.\n\n9IÖÆ×÷³§Ãâ·Ñ Statistics Seminar schedule: https://mcgillstat.g
 ithub.io/\n\n \n\nhttps://mcgill.zoom.us/j/83436686293?pwd=b0RmWmlXRXE3OWR
 6NlNIcWF5d0dJQT09\n\nMeeting ID: 834 3668 6293\n\nPasscode: 12345\n\n \n\n
 \n	\n		 \n	\n\n
DTSTART:20220325T193000Z
DTEND:20220325T203000Z
SUMMARY:Rina Foygel Barber (University of Chicago)
URL:/mathstat/channels/event/rina-foygel-barber-univer
 sity-chicago-338696
END:VEVENT
END:VCALENDAR