Abstract

In this paper particle filters for dynamic state space models handling unknown static parameters are discussed. The approach is based on marginalizing the static parameters out of the posterior distribution such that only the state vector needs to be considered. Such a marginalization can always be applied. However, real-time applications are only possible when the distribution of the unknown parameters given both observations and the hidden state vector depends on some low-dimensional sufficient statistics. Such sufficient statistics are present in many of the commonly used state space models. Marginalizing the static parameters avoids the problem of impoverishment which typically occur when static parameters are included as part of the state vector. The filters are tested on several different models, with promising results.

Citations

911	Condensation: conditional density propagation for visual tracking - Isard - 1998
751	Novel Approach to Nonlinear/Non-Gaussian Bayesian State Estimation - Gordon, Salmond, et al. - 1993
475	Optimal Filtering - Anderson, Moore - 1979
463	On Sequential Monte Carlo Sampling Methods for Bayesian Filtering - Doucet, Godsill, et al. - 2000
380	Forecasting, Structural Time Series Models and the Kalman Filter - Harvey - 1989
360	Filtering via simulation: Auxiliary particle filter - Pitt, Shephard - 1999
339	Sequential Monte Carlo methods for dynamic systems - Liu, Chen - 1998
276	Monte Carlo filter and smoother for non-Gaussian nonlinear state space models - Kitagawa - 1996
243	CONDENSATION—Conditional Density Propagation for Visual Tracking - Isard, Blake - 1998
169	On sequential simulation-based methods for Bayesian filtering - Doucet - 1998
153	Statistical aspects of ARCH and stochastic volatility. Time series models in econometrics, finance and other fields - Shephard - 1996
132	Sequential imputations and Bayesian missing data problems - Kong, Liu, et al. - 1994
96	2001: An introduction to sequential Monte Carlo methods. Sequential Monte Carlo Methods in Practice - Doucet, Freitas, et al.
94	Following A Moving Target–Monte Carlo Inference for Dynamic Bayesian Models - Gilks, Berzuini - 2001
92	Combined Parameter and State Estimation in Simulation-Based Filtering,” in Sequential Monte Carlo Methods in Practice - Liu, West - 2001
86	Bayesian Forecasting and Dynamic Models, 2nd edn - West, Harrison - 1997
85	Data assimilation in meteorology and oceanography - Ghil, Malanotte-Rizzoli - 1991
73	Partial non-Gaussian state space - Shephard - 1994
57	Dynamic conditional independence models and Markov chain Monte Carlo methods - Berzuini, Best, et al. - 1997
45	Dynamic Generalized Linear Models and Bayesian Forecasting - West, Harrison, et al. - 1985
37	Logvinenko A theoretical framework for sequential importance sampling and resampling - Lui, Chen, et al. - 2001
33	Mixture models, Monte Carlo, Bayesian updating and dynamic models - West - 1993
32	A self-organising state-space model - Kitagawa - 1998
26	Sequential Monte Carlo in practice - Doucet, Freitas, et al. - 2001
23	Building robust simulation-based filters for evolving data sets - Carpenter, Clifford, et al. - 1999
22	Markov chain Monte Carlo for dynamic generalised linear models - Gamerman - 1998
21	Monte Carlo and smoother for non-Gaussian nonlinear state space models - Kitagawa - 1996
18	Hierarchical Bayesian space–time models - Wikle, Berliner, et al. - 1998
18	BUGS 0.5, Bayesian Inference Using Gibbs Sampling Manual (Version ii). Cambridge, UK:Medical Research Council Biostatistics Unit. Available: http://www.mrc-bsu.cam.ac.uk/bugs/ documentation/content.shtml [accessed 26 - Spiegelhalter, Thomas, et al. - 1996
15	Bayesian forecasting and dynamic models SpringerVerlag Inc - West, Harrison - 1997
14	Fixed-Lag Smoothing Using Sequential Importance Sampling - Clapp, Godsill - 1999
10	Monte Carlo integration in general dynamic models - Muller - 1991
10	Hierarchical Bayesian space-time models. Environmental and Ecological Statistics - Wikle, Berliner, et al. - 1998
8	Monte Carlo Filters for Non-linear State Estimation,” Automatica - Bølviken, Acklam, et al. - 2001
7	State space and hidden Markov models, in Complex stochastic systems - Künsch - 2001
5	Building robust simulation-based for evolving data sets - Carpenter, Cliord, et al. - 1999
5	Deterministic and stochastic particle filters in state space models - Bolviken, Storvik - 2000
3	A longitudal study of emergence room visits and air pollution for Prince George British Columbia - Jrgensen, Lundbye-Christensen, et al. - 1996
1	State space and hidden Markov models," in Complex Stochastic Systems - Kunsch - 2001
1	Optimal Electrical Engineering Series - Anderson, Moore - 1979
1	Strdal, \Monte Carlo for non-linear state estimation - Blviken, Acklam, et al. - 1997
1	Discussion on Monte Carlo methods for dynamic Bayesian problems by Berzuini and Gilks," Discussion contribution to the HSSS workshop on Models and inference in HSSS: Recent developments and perspectives - Storvik - 2001
1	Deterministic and stochastic particle in state space models," in Sequential Monte Carlo in Practice - Blviken, Storvik - 2001
1	Monte Carlo integration in general dynamicmodels - Muller - 1991
1	Dynamic conditional independencemodels and Markov chain Monte Carlo methods - Berzuini, Best, et al. - 1997
1	Some further topics on Monte Carlo methods for dynamic Bayesian problems - Storvik
1	Deterministic and stochastic particle lters in state space models - Blviken, Storvik - 2001