BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Learning under model misspecification - Andrés R. Masegosa\, Univ
 ersidad de Almería (Spain)
DTSTART:20210202T131500Z
DTEND:20210202T141500Z
UID:TALK156022@talks.cam.ac.uk
CONTACT:Mateja Jamnik
DESCRIPTION:"Join us on Zoom":https://zoom.us/j/99166955895?pwd=SzI0M3pMVE
 kvNmw3Q0dqNDVRalZvdz09\n\nBayesian statistics is one of the most employed 
 tools for uncertainty modelling in deep learning. But\, in this talk\, we 
 will present recent research which cast doubts about the optimality of the
  Bayesian approach for this task. More precisely\, we will present a novel
  PAC-Bayesian analysis of Bayesian model averaging showing that it is only
  optimal for generalization when the model class is perfectly specified\, 
 something which rarely happens in practice. \n\nBy building on this theore
 tical analysis\, we will introduce a novel learning framework based on the
  minimization of a new family of PAC-Bayesian bounds which explicitly assu
 me that the model class is misspecified (a much more realistic assumption)
 . We will also discuss strong connections with deep ensemble methods. 
LOCATION:Zoom
END:VEVENT
END:VCALENDAR