 |
 |
Statistics, Applied Probability & Operational Research Seminar Date: 13 October 2011 Time: 2:15 p.m. Venue: Lecture Theatre, Department of Statistics, University of Oxford, 1 South Parks Road, Oxford, OX1 3TG Speaker: Nick Whiteley<http://maths.bris.ac.uk/%7Emanpw/>,<http://maths.bris.ac.uk/%7Emanpw/> School of Mathematics, University of Bristol Title: Stability properties of some particle filters Abstract: Particle filtering has become a very popular method for inferential computation in state space models. What arguably makes particle filters useful is that they exhibit some form of stability: it is commonly observed that the stochastic errors associated with the particle approximation do not explode over time, so that these algorithms can be applied successfully at a computational cost which scales slowly with the length of the data record under analysis. Whilst there is now a rich literature on various theoretical properties of particle filters, most existing results which explain their stability properties rely on strong mixing assumptions which do not hold in typical applications. One of the main obstacles is dealing with a non-compact state space. This talk will describe recent developments which allow some stability properties of particle filters to be verified when the state space is non-compact. This involves some structural connections to spectral properties of linear operators arising from the state space model and the Feynman-Kac formulae which underlie the particle algorithm. Tea, Coffee and biscuits are served after the Seminar in the Statistics Common Room at 3.30pm The webpage for the seminar: http://www.stats.ox.ac.uk/news_and_events/weekly_seminars Christine ------------ Christine Stone Administrative and Personal Assistant Department of Statistics, University of Oxford 1 South Parks Road Oxford OX1 3TG Tel: +44 (0)1865 272866/60 Fax: +44 (0)1865 272595 P Please consider the environment - do you really need to print this email? You may leave the list at any time by sending the command SIGNOFF allstat to [log in to unmask], leaving the subject line blank.