Information for Staff
Select dates to view past and future events
13:00 - 14:00 5 October 2012
Bayesian Parameter Inference in State-Space Models using Particle Markov chain Monte Car
Roberts G06 Sir Ambrose Fleming LT |
Malet Place | London | WC1E 7JE | United Kingdom
Admission: Free of charge
Arnaud Doucet, Professor of Statistics, University of Oxford, Arnaud Doucet obtained his PhD Degree from University Paris XI in 1997. He has held previously faculty positions at the University of Melbourne, the University of Cambridge, the Institute of Statistical Mathematics in Tokyo and was a Canada Research Chair at the University of British Columbia. He joined the Department of Statistics of the University of Oxford in 2011 where he is currently Professor. He is Associate editor of the Annals of Statistics and ACM Transactions on Modeling and Computer Simulation. His research areas include Monte Carlo methods, Bayesian statistics, dynamic models and their applications
Standard Markov chain Monte Carlo (MCMC) methods to perform Bayesian inference for both states and parameter in state-space models can be very inefficient and/or not even applicable for complex models.
I will discuss how it is possible to come up with a new class of efficient MCMC algorithms using particle filtering proposals. This yields the class of particle MCMC methods.
One crucial practical problem for particle MCMC is the selection of the number of particles for the proposal. Essentially, a trade-off is needed. If too many particles are used, then the particle MCMC scheme has similar properties to the case where the likelihood of the parameter is exactly known but will be expensive. If too few particles are used, this is at the expense of slower mixing in the resulting Markov chain.
I will address how one should select the number of particles so as to maximize the efficiency of the particle MCMC scheme.
020 7679 0481 | firstname.lastname@example.org
Get directions to the venue from anywhere in the UK
Powered by Google