Importance Sampling Applied to Pincus Maximization for Particle Filter MAP Estimation
2012 (English)In: 15th International Conference on Information Fusion (FUSION), 2012, Proceeding, International Society of Information Fusion (ISIF) , 2012, , 7 p.114-120 p.Conference paper (Refereed)
Sequential Monte Carlo (SMC), or Particle Filters(PF), approximate the posterior distribution in nonlinear ﬁlteringarbitrarily well, but the problem how to compute a state estimateis not always straightforward. For multimodal posteriors, themaximum a posteriori (MAP) estimate is a logical choice, butit is not readily available from the SMC output. In principle,the MAP can be obtained by maximizing the posterior density obtained e.g. by the particle based approximation of theChapman-Kolmogorov equation. However, this posterior is amixture distribution with many local maxima, which makes theoptimization problem very hard. We suggest an algorithm forestimating the MAP using the global optimization principle ofPincus and subsequently outline the frameworks for estimatingthe ﬁlter and marginal smoother MAP of a dynamical systemfrom the SMC output.
Place, publisher, year, edition, pages
International Society of Information Fusion (ISIF) , 2012. , 7 p.114-120 p.
particle ﬁlter, particle smoother, maximum a posteriori, global optimization
IdentifiersURN: urn:nbn:se:liu:diva-79585ISBN: 978-0-9824438-4-2 (print)ISBN: 978-1-4673-0044-5 (online)OAI: oai:DiVA.org:liu-79585DiVA: diva2:543905
15th International Conference on Information Fusion (FUSION), 2012, July 9-12, Singapore
FundereLLIIT - The Linköping‐Lund Initiative on IT and Mobile CommunicationsSwedish Research Council