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:5519@i2m.univ-amu.fr
DTSTART;TZID=Europe/Paris:20131011T140000
DTEND;TZID=Europe/Paris:20131011T150000
DTSTAMP:20241030T092152Z
URL:https://www.i2m.univ-amu.fr/evenements/s-loustau-larema-univ-angers-at
 -frumam-inverse-statistical-learning-from-minimax-to-algorithm/
SUMMARY: (...): S. Loustau (LAREMA\, Univ. Angers) at Frumam : Inverse Stat
 istical Learning - From minimax to algorithm
DESCRIPTION:: Inverse Statistical Learning : From minimax to algorithm\n\nB
 y Sébastien Loustau\\\, LAREMA\\\, Univ. Angers.\n\nWe propose to conside
 r the problem of statistical learning when we observe a contaminated sampl
 e. More precisely\\\, we state minimax rates of convergence in classificat
 ion with errors in variables for deconvolution empirical risk minimizers. 
 These fast rates depends on the ill-posedness\\\, the margin and the compl
 exity of the problem. The cornerstone of the proof is a bias variance deco
 mposition of the excess risk.\nAfter a theoretical study of the problem\\\
 , we turn out into more practical considerations by presenting a new algor
 ithm for noisy finite dimensional clustering called noisy K-means. The alg
 orithm is based on a two-step procedure : a deconvolution step to deal wit
 h noisy inputs and Newton's iteration as the popular k-means.
CATEGORIES:Séminaire,Statistique
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/Paris
X-LIC-LOCATION:Europe/Paris
BEGIN:DAYLIGHT
DTSTART:20130331T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
END:DAYLIGHT
END:VTIMEZONE
END:VCALENDAR