BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.2.3.1//EN
TZID:Europe/Paris
X-WR-TIMEZONE:Europe/Paris
BEGIN:VEVENT
UID:5526@i2m.univ-amu.fr
DTSTART;TZID=Europe/Paris:20131128T160000
DTEND;TZID=Europe/Paris:20131128T170000
DTSTAMP:20241030T093324Z
URL:https://www.i2m.univ-amu.fr/evenements/p-machart-inria-rennes-at-cmi-r
 econciling-priors-and-priors-without-prejudice/
SUMMARY: (...): P. Machart (Inria Rennes) at CMI : Reconciling "priors" and
  "priors" without prejudice?
DESCRIPTION:: Reconciling "priors" and "priors" without prejudice?\n\nBy Pi
 erre Machart\\\, Inria\\\, Rennes.\n\nThere are two major routes to addres
 s linear inverse problems. Whereas regularization-based approaches build e
 stimators as solutions of penalized regression optimization problems\\\, B
 ayesian estimators rely on the posterior distribution of the unknown\\\, g
 iven some assumed family of priors. While these may seem radically differe
 nt approaches\\\, recent results have shown that\\\, in the context of add
 itive white Gaussian denoising\\\, the Bayesian conditional mean estimator
  is always the solution of a penalized regression problem. The contributio
 n of this paper is twofold. First\\\, we extend the additive white Gaussia
 n denoising results to general linear inverse problems with colored Gaussi
 an noise. Second\\\, we characterize conditions under which the penalty fu
 nction associated to the conditional mean estimator can satisfy certain po
 pular properties such as convexity\\\, separability\\\, and smoothness. Th
 is sheds light on some tradeoff between computational efficiency and estim
 ation accuracy in sparse regularization\\\, and draws some connections bet
 ween Bayesian estimation and proximal optimization.
CATEGORIES:Séminaire,Statistique
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/Paris
X-LIC-LOCATION:Europe/Paris
BEGIN:STANDARD
DTSTART:20131027T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
END:STANDARD
END:VTIMEZONE
END:VCALENDAR