BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Virtual Seminar: “Delayed-acceptance Sequential Monte Carlo” -
  Joshua Bon\, Queensland University of Technology
DTSTART:20200521T090000Z
DTEND:20200521T100000Z
UID:TALK141661@talks.cam.ac.uk
CONTACT:Alison Quenault
DESCRIPTION:Delayed-acceptance is a technique for reducing computational e
 ffort for Bayesian models with expensive likelihoods. Using delayed-accept
 ance kernels in MCMC can reduce the number of expensive likelihoods evalua
 tions required to approximate a posterior expectation to a given accuracy.
  It uses a surrogate\, or approximate\, likelihood to avoid evaluation of 
 the expensive likelihood when possible. Importantly\, delayed-acceptance k
 ernels preserve the intended targeted distribution of the Markov chain\, w
 hen viewed as an extension of a Metropolis-Hastings kernel. Within the seq
 uential Monte Carlo (SMC) framework\, we utilise the history of the sample
 r to adaptively tune the surrogate likelihood to yield better approximatio
 ns of the expensive likelihood\, and use a surrogate first annealing sched
 ule to further increase computational efficiency. Moreover\, we propose a 
 framework for optimising computation time whilst avoiding particles degene
 racy\, which encapsulates existing strategies in the literature. Overall\,
  we develop a novel algorithm for computationally efficient SMC with expen
 sive likelihood functions. The method is applied to static Bayesian models
 \, which we demonstrate on toy and real examples.
LOCATION:Virtual Seminar 
END:VEVENT
END:VCALENDAR