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:Statistics Reading Group
SUMMARY:Reversible jump MCMC - Eleni Bakra\, MRC Biostatis
tics Unit
DTSTART;TZID=Europe/London:20100203T163000
DTEND;TZID=Europe/London:20100203T180000
UID:TALK23100AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/23100
DESCRIPTION:Reversible jump Markov chain Monte Carlo (RJMCMC)
is an extension of the\nMetropolis-Hastings algori
thm for stationary distributions of variable\ndime
nsion. It has been successfully applied in a wide
variety of settings\,\nincluding the challenging p
roblem of determining the number of components in\
na finite mixture and determining the number of st
ates in a hidden Markov\nmodel. In this talk\, RJM
CMC is considered as an approach to Bayesian model
\nselection problems and for this reason\, it is u
sed to explore the sampling\nspace that consists o
f several models of different dimension. The RJMCM
C\nalgorithm usually considers a selection of move
types\, some of which explore\nthe parameter spac
e within a model\, and others which propose change
s to the\ndimensionality of the model. The choice
of the proposal mechanism is crucial\nto the perfo
rmance of the algorithm. Therefore\, several metho
ds have been\nproposed in the literature on how to
choose the proposal mechanism of the\nalgorithm.
In this talk\, I will describe some of these metho
ds and I will\nalso introduce a new one.\n\nThe or
iginal reversible jump MCMC paper can be found "he
re":http://www.jstor.org/stable/pdfplus/2985194.pd
f
LOCATION:MR5\, CMS
CONTACT:Richard Samworth
END:VEVENT
END:VCALENDAR