BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Machine Learning Reading Group @ CUED
SUMMARY:Exchangeability - Peter Orbanz (University of Camb
ridge)
DTSTART;TZID=Europe/London:20110317T140000
DTEND;TZID=Europe/London:20110317T153000
UID:TALK29126AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/29126
DESCRIPTION:Data is called exchangeable if the probability of
a sample depends only on the values that are obser
ved\, but not on the order in which they occur. Th
is simple assumption has a rather surprising cons
equence: An\nexchangeable distribution can be deco
mposed completely into a shared\, underlying rando
m "pattern" and a component representing independe
nt "randomness" in each observation. This result i
s de Finetti's theorem. I will devote roughly the
first half of the talk to the theorem and its imme
diate implications for Bayesian statistics. In the
remaining time\, I hope to provide a glimpse of m
ore recent results and of the fascinating picture
which emerges with them\, which relates symmetry a
nd invariance principles to the design of probabil
ity models.\n
LOCATION:Engineering Department\, CBL Room 438
CONTACT:Konstantina Palla
END:VEVENT
END:VCALENDAR