BEGIN:VCALENDAR
VERSION:2.0
PRODID:researchseminars.org
CALSCALE:GREGORIAN
X-WR-CALNAME:researchseminars.org
BEGIN:VEVENT
SUMMARY:LI Cheuk Ting (Chinese University of Hong Kong)
DTSTART:20231011T150000Z
DTEND:20231011T161500Z
DTSTAMP:20260423T021420Z
UID:AAIT/22
DESCRIPTION:Title: <a href="https://researchseminars.org/talk/AAIT/22/">Fu
 nctional Representation Lemma and Minimum Entropy Coupling</a>\nby LI Cheu
 k Ting (Chinese University of Hong Kong) as part of Seminar on Algorithmic
  Aspects of Information Theory\n\n\nAbstract\nThe functional representatio
 n lemma states that for two jointly-distributed random variables X\, Y\, i
 t is possible to express Y as a function of X and another random variable 
 Z that is independent of X. There are two interesting extreme points. If w
 e minimize H(Y|Z)\, the minimum can be bounded by the "strong functional r
 epresentation lemma"\, and has implications in lossy compression and chann
 el simulation. If we instead minimize H(Z)\, this is equivalent to the min
 imum entropy coupling problem\, which concerns finding the coupling of sev
 eral given distributions that gives the smallest joint entropy. In this ta
 lk\, we will discuss several recent developments regarding these problems.
 \n
LOCATION:https://researchseminars.org/talk/AAIT/22/
END:VEVENT
END:VCALENDAR