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:Signal Processing and Communications Lab Seminars
SUMMARY:On the convergence of Adaptive sequential Monte Ca
rlo Methods - Dr Ajay Jasra\, Department of Statis
tics and Applied Probability\, National University
of Singapore
DTSTART;TZID=Europe/London:20130603T140000
DTEND;TZID=Europe/London:20130603T150000
UID:TALK45523AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/45523
DESCRIPTION:In several implementations of Sequential Monte Car
lo (SMC) methods\, it is natural and important in
terms of algorithmic efficiency\, to exploit the i
nformation on the history of the particles to opti
mally tune their subsequent propagations. In the f
ollowing talk we provide an asymptotic theory for
a class of such adaptive SMC methods. Our theoreti
cal framework developed here will cover for instan
ce\, under assumptions\, the algorithms in Chopin
(2002)\, Jasra et al (2011)\, Schafer & Chopin (20
13). There are limited results about the theoretic
al underpinning of such adaptive methods: we will
bridge this gap by providing a weak law of large n
umbers (WLLN) and a central limit theorem (CLT) fo
r some of the algorithms. The latter seems to be t
he first result of its kind in the literature and
provides a formal justification of algorithms that
are used in many practical scenarios. This is a j
oint work with Alex Beskos (NUS/UCL).
LOCATION:LR11\, Engineering\, Department of
CONTACT:Dr Ramji Venkataramanan
END:VEVENT
END:VCALENDAR