BEGIN:VCALENDAR
PRODID:-//Columba Systems Ltd//NONSGML CPNG/SpringViewer/ICal Output/3.3-
 M3//EN
VERSION:2.0
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
DTSTAMP:20211022T123653Z
DTSTART:20211117T140000Z
DTEND:20211117T150000Z
SUMMARY:Kolyan Ray  - Variational Bayes for high-dimensional linear regre
 ssion with sparse priors
UID:{http://www.columbasystems.com/customers/uom/gpp/eventid/}ckt-ku8agzg
 9-n8nta0
DESCRIPTION:Kolyan Ray\, Lecturer in Statistics at the Department of Math
 ematics\, Imperial College London is our speaker for the Statistics semi
 nar series.\n\nTitle: Variational Bayes for high-dimensional linear regr
 ession with sparse priors\n\nAbstract: A core problem in Bayesian statis
 tics is approximating difficult to compute posterior distributions. In v
 ariational Bayes (VB)\, a method from machine learning\, one approximate
 s the posterior through optimization\, which is typically faster than Ma
 rkov chain Monte Carlo. We study a mean-field (i.e. factorizable) VB app
 roximation to Bayesian model selection priors\, including the popular sp
 ike-and-slab prior\, in sparse high-dimensional linear regression. We es
 tablish convergence rates for this VB approach\, studying conditions und
 er which it provides good estimation. We also discuss computational issu
 es and study the empirical performance of the algorithm.
STATUS:TENTATIVE
TRANSP:TRANSPARENT
CLASS:PUBLIC
LOCATION:Zoom link: https://zoom.us/j/92947173491
END:VEVENT
END:VCALENDAR