For a class of non-uniformly ergodic partly observable Markov processes, under observations subject to a Wiener process or i.i.d. noise, it is shown that a wrong initial data is forgotten with a certain rate.
On démontre que pour une classe de processus de Markov non-uniformément ergodiques avec des observations perturbées par un mouvement Brownien, le fait d'avoir des données initiales erronées est asymptotiquement oublié. La vitesse de convergence est explicitée.
Accepted:
Published online:
Marina L. Kleptsyna 1, 2; Alexander Yu. Veretennikov 2, 3
@article{CRMATH_2007__344_11_727_0, author = {Marina L. Kleptsyna and Alexander Yu. Veretennikov}, title = {On ergodic filters with wrong initial data}, journal = {Comptes Rendus. Math\'ematique}, pages = {727--731}, publisher = {Elsevier}, volume = {344}, number = {11}, year = {2007}, doi = {10.1016/j.crma.2007.04.015}, language = {en}, }
Marina L. Kleptsyna; Alexander Yu. Veretennikov. On ergodic filters with wrong initial data. Comptes Rendus. Mathématique, Volume 344 (2007) no. 11, pp. 727-731. doi : 10.1016/j.crma.2007.04.015. https://comptes-rendus.academie-sciences.fr/mathematique/articles/10.1016/j.crma.2007.04.015/
[1] Exponential stability for nonlinear filtering of diffusion processes in a noncompact domain, Ann. Probab., Volume 26 (1998) no. 4, pp. 1552-1574
[2] Exponential stability for nonlinear filtering, Ann. Inst. H. Poincaré Probab. Statist., Volume 33 (1997) no. 6, pp. 697-725
[3] Asymptotic stability of the Wonham filter: ergodic and nonergodic signals, SIAM J. Control Optim., Volume 43 (2004) no. 2, pp. 643-669 (electronic)
[4] Exponential stability in discrete-time filtering for bounded observation noise, Systems Control Lett., Volume 30 (1997), pp. 185-193
[5] On discrete time filtres with wrong initial data (revised version), 2006 http://www.univ-lemans.fr/sciences/statist/download/Kleptsyna/filt19fg.pdf (Prépublication no 06-3, de LSP, Univ. du Maine)
[6] On continuous time filtres with wrong initial data, 2006 http://www.univ-lemans.fr/sciences/statist/download/Kleptsyna/filt36c.pdf (Prépublication no 06-2 de LSP, Univ. du Maine)
[7] A property of the solutions of parabolic equations with measurable coefficients, Izv. Akad. Nauk SSSR Ser. Mat., Volume 44 (1980) no. 1, pp. 161-175 239 (in Russian)
[8] A robustification approach to stability and to uniform particle approximation of nonlinear filters: the example of pseudo-mixing signals, Stochastic Process. Appl., Volume 106 (2003), pp. 279-316
[9] Probability Measures on Metric Spaces, Academic Press, New York and London, 1967
[10] Stability of the pathwise filter equations for a time dependent signal on , Appl. Math. Optim., Volume 52 (2005), pp. 39-71
[11] On polynomial mixing and the rate of convergence for stochastic differential and difference equations, Teor. Veroyatn. i Primenen., Volume 44 (1999) no. 2, pp. 312-327 (Engl. transl. in Theory Probab. Appl., 44, 2, 2000, pp. 361-374)
Cited by Sources:
Comments - Policy