Jing Lei and Peter Bickel (2013). On convergence of recursive Monte Carlo filters in non-compact state spaces. Vol. 23, No. 1, 429-450.

Statistica Sinica 23 (2013), 429-450

doi:http://dx.doi.org/10.5705/ss.2011.187

ON CONVERGENCE OF RECURSIVE MONTE CARLO

FILTERS IN NON-COMPACT STATE SPACES

Jing Lei and Peter Bickel

Carnegie Mellon University and University of California, Berkeley

Abstract: We consider the particle filter approximation of the optimal filter in non-compact state space models. A time-uniform convergence result is built on top of a filter stability argument developed by [#!DoucMR09!#], under the assumption of a heavy-tailed state process and an informative observation model. We show that an existing set of sufficient conditions for filter stability is also sufficient, with minor modifications, for particle filter convergence. The rate of convergence is also given and depends on both the sample size and the tail behavior of the transition kernel.

Key words and phrases: Consistency, particle filter, state space model.