Abstract
In this paper we study ergodic properties of hidden Markov models with a generalized observation structure. In particular sufficient conditions for the existence of a unique invariant measure for the pair filter-observation are given. Furthermore, necessary and sufficient conditions for the existence of a unique invariant measure of the triple state-observation-filter are provided in terms of asymptotic stability in probability of incorrectly initialized filters. We also study the asymptotic properties of the filter and of the state estimator based on the observations as well as on the knowledge of the initial state. Their connection with minimal and maximal invariant measures is also studied.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Atar R, Zeitouni O (1997) Lyapunov exponents for finite-state nonlinear filtering. SIAM J Control Optimiz 35:36–55
Baxendale P, Chigansky P, Liptser R (2004) Asymptotic stability of the Wonham filter: ergodic and nonergodic signals. SIAM J Control Optimiz 43:643–669
Bielecki T, Hernandez-Hernandez D, Pliska SR (1999) Risk-sensitive control of finite-state Markov Chains in discrete time, with applications to portfolio management. Math Meth Oper Res 50:167–188
Bielecki TR, Pliska SR (1999) Risk-sensitive dynamic asset management. JAMO 39:337–360
Borkar VS (1998) Ergodic control of partially observed Markov chains. Sys Control Lett 34:185–189
Budhiraja A (2001) Ergodic properties of the nonlinear filter. Stoch Proc Appl 95:1–24
Budhiraja A (2003) Asymptotic stability, ergodicity and other asymptotic properties of the nonlinear filter. Ann IH Poincare PR 39,6:919–941
Budhiraja A, Kushner HJ (2001) Monte Carlo algorithms and asymptotic problems in nonlinear filtering. In Stochastics in finite/infinite dimensions, Birkhäuser
Douc R, Matias C (2001) Asymptotics of the maximum likelihood estimator for general hidden Markov models. Bernoulli 7:381–420
Ephraim Y, Merhav N (2002) Hidden Markov Processes. IEEE Trans Inf Th 48:1518–1569
Ikeda N, Watanabe S (1981) Stochastic differential equations and diffusion processes. North-Holland, Amsterdam
Kaijser T (1975) A limit theorem for partially observed Markov chains. Annals Prob 3:677–696
Khasminskii RZ (1980) Stochastic Stability of Differential Equations. Sijthoff & Noordhoff, Netherlands
Kunita H (1971) Asymptotic behaviour of the nonlinear filtering errors of Markov process. J Multivariate Anal 1:365–393
Kunita H (1991) Ergodic properties of nonlinear filtering processes. In: Aleksander KC, Watkins JC (eds) Spatial Stochastic Processes. Progr Probab 19, 233–256, Birkhäuser
Le Gland F, Mevel L (2000) Exponential forgetting and geometric ergodicity in hidden Markov models. Math Control Sign Sys 13:63–93
Liverani C (1995) t Decay of correlations. Ann Math 142:239–301
Meyn SP, Tweedie RL (1993) Markov chains and stochastic stability. Springer, Berlin Heidelberg, New York
Ocone D, Pardoux E (1996) Asymptotic stability of the optimal filter with respect to its initial conditions. SIAM J Control Optimiz 34:226–243
Runggaldier WJ, Stettner Ł (1994) Approximations of discrete-time partially observed control problems. Giardini Editori, Pisa
Stettner Ł (1989) On invariant measures of filtering processes. In: Helmes K, Christopeit N, Kohlmann M (eds) Stochastic differential systems, Lecture Notes in Control and Inf Scifs 126, Springer, Berlin Heidelberg, New York, pp. 279–292
Author information
Authors and Affiliations
Corresponding author
Additional information
Work partially supported by grants MIUR-PRIN 2001, PBZ KBN 016/P03/99 and IMPAN-BC Centre of Excellence
Rights and permissions
About this article
Cite this article
Di Masi, G., Stettner, Ł. Ergodicity of hidden Markov models. Math. Control Signals Syst. 17, 269–296 (2005). https://doi.org/10.1007/s00498-005-0153-8
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00498-005-0153-8