Moderate deviations for particle filtering

DOI10.1214/105051604000000657MaRDI QIDQ1774189zbMATH OpenOpenAlexFDO

Authors Randal Douc, Arnaud Guillin, Jamal Najim

Publication date 29 April 2005

Published in The Annals of Applied Probability (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/math/0401058

Large deviations (60F10) Signal detection and filtering (aspects of stochastic processes) (60G35) Filtering in stochastic control theory (93E11)

Abstract: Consider the state space model (X_t,Y_t), where (X_t) is a Markov chain, and (Y_t) are the observations. In order to solve the so-called filtering problem, one has to compute L(X_t|Y_1,...,Y_t), the law of X_t given the observations (Y_1,...,Y_t). The particle filtering method gives an approximation of the law L(X_t|Y_1,...,Y_t) by an empirical measure frac{1}{n}sum_1^ndelta_{x_{i,t}}. In this paper we establish the moderate deviation principle for the empirical mean frac{1}{n}sum_1^npsi(x_{i,t}) (centered and properly rescaled) when the number of particles grows to infinity, enhancing the central limit theorem. Several extensions and examples are also studied.

Review text

A ``state space in discrete time model underlies the paper being reviewed. The state is a Markov chain (\({\underline{X}}_n\in \mathbb R^p\)) and the observations (\({\underline{Y}}_n\in \mathbb R^q\)) are independent given the current state. The conditional law of the current state given the preceding one and that of the current observation given the current state are each assumed to have a conditional density with respect to an a priori law. The problem is that of estimating recursively the law of the current state given current and past observations. It is usually impossible to obtain that law analytically and particle filters are means to obtain approximations of that law in the form of an average of point masses. They typically sample from a recursively defined conditional density for the state and are related to genetic algorithms. The theory then covers questions of convergence of these particle filters. The asymptotic behaviour of averages of centered and scaled functional values of the states is investigated and bounds for sequences of the following type are provided: \(\frac{1}{b_N^2}\ln [P_T(M_N^T\in \Gamma)]\), where \(M_N^T\) is the average considered, \(T\) the observation span, \(N\) the sampling size, \(\Gamma\) a Borel set and \(b_N\) a sequence increasing more slowly than \(\sqrt{N}.\) \(P_T\) represents the conditional law. The bounds are in terms of the behaviour of a ``good rate function on the interior and closure of \(\Gamma.\) This rate function depends on the asymptotic covariance matrix of the considered functionals of the state. Towards the end of the paper three examples are considered, but only as far as to show that the assumptions obtain indeed that are required for the results of the paper to be valid. Given the complexity of particle filters constructions, extended applications in a ``real world context would seem necessary for the evaluation of their scope.

Recommendations

Cites work

Cited in

(4)

This page was built for publication: Moderate deviations for particle filtering

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1774189)