BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Gaussian concentration bounds and finitary coding
DTSTART:20250326T161500
DTEND:20250326T171500
DTSTAMP:20260531T193258Z
UID:abf2c81398f1ebe88c4e8eab7e3988b2b1efb2000ea230e4d46959d3
CATEGORIES:Conferences - Seminars
DESCRIPTION:Sandro Gallo\, Universide Federal de São Carlos (Brazil)\n\n\
 nThe study of concentration inequalities focuses on upper bounds for the p
 robability that certain statistics of (fixed-size) random samples deviate 
 significantly from their mean (or median). For i.i.d. samples\, what we re
 fer to as a "Gaussian concentration bound" is a specific case of a concent
 ration inequality\, commonly known in the literature as McDiarmid’s ineq
 uality. More broadly\, such bounds are expected to hold for well-behaved s
 tatistics (e.g.\, Lipschitz continuous functions) and for samples of weakl
 y dependent random variables. In this talk I will relate the occurrence of
  such bounds to the concept of finitary coding (or factor) coming from dyn
 amical systems. As a consequence\, I will present recent results establish
 ing gaussian concentration bounds for a wide class of random fields on $\\
 mathds{Z}^d$\, in particular the Ising model above the critical temperatur
 e in any dimension.\n \nThis presentation is based on joint work with Jea
 n-René Chazottes (CNRS & École Polytechnique\, Palaiseau) and Daniel Y. 
 Takahashi (Instituto do Cérebro\, UFRN\, Brazil).\n\n-- A Probability and
  Stochastic Analysis Seminar--
LOCATION:C 1 517 https://plan.epfl.ch/?room=%3DGR%20C1%20517&dim_floor=1&l
 ang=en&dim_lang=en&tree_groups=centres_nevralgiques_grp%2Cmobilite_acces_g
 rp%2Crestauration_et_commerces_grp%2Censeignement%2Cservices_campus_g
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR