Adaptive stopping for fast particle smoothing
2013 (English)In: Proceedings of the 38th International Conference on Acoustics, Speech, and Signal Processing (ICASSP), IEEE , 2013, 6293-6297 p.Conference paper (Refereed)
Particle smoothing is useful for offline state inference and parameter learning in nonlinear/non-Gaussian state-space models. However, many particle smoothers, such as the popular forward filter/backward simulator (FFBS), are plagued by a quadratic computational complexity in the number of particles. One approach to tackle this issue is to use rejection-sampling-based FFBS (RS-FFBS), which asymptotically reaches linear complexity. In practice, however, the constants can be quite large and the actual gain in computational time limited. In this contribution, we develop a hybrid method, governed by an adaptive stopping rule, in order to exploit the benefits, but avoid the drawbacks, of RS-FFBS. The resulting particle smoother is shown in a simulation study to be considerably more computationally efficient than both FFBS and RS-FFBS.
Place, publisher, year, edition, pages
IEEE , 2013. 6293-6297 p.
Sequential Monte Carlo, particle smoothing, backward simulation
Signal Processing Control Engineering
IdentifiersURN: urn:nbn:se:liu:diva-93461DOI: 10.1109/ICASSP.2013.6638876ISBN: 978-147990356-6OAI: oai:DiVA.org:liu-93461DiVA: diva2:624920
38th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013; Vancouver, BC; Canada
FunderSwedish Research Council, 621-2010-5876