Online sequential Monte Carlo EM algorithm

Author

Cappé, Olivier

Author_Institution

LTCI, Telecom ParisTech, Paris, France

fYear

2009

Firstpage

37

Lastpage

40

Abstract

Online (or recursive) estimation of fixed model parameters in general state-space models is a crucial but often difficult task. This paper is about likelihood-based point estimation, showing that an online EM (Expectation-Maximization) algorithm recently proposed for discrete hidden Markov models can be extended to more general settings, including non-linear non-Gaussian state-space models that necessitate the use of sequential Monte Carlo filtering approximations. The performance of the proposed online sequential Monte Carlo EM algorithm is illustrated on numerical examples.

Keywords

Monte Carlo methods; approximation theory; expectation-maximisation algorithm; filtering theory; hidden Markov models; parameter estimation; discrete hidden Markov model; expectation-maximization algorithm; filtering approximation; online sequential Monte Carlo algorithm; parameter estimation; state-space model; Filtering; Hidden Markov models; Monte Carlo methods; Parameter estimation; Recursive estimation; Signal processing algorithms; Sliding mode control; Smoothing methods; State estimation; Tin;

fLanguage

English

Publisher

ieee

Conference_Titel

Statistical Signal Processing, 2009. SSP '09. IEEE/SP 15th Workshop on

Conference_Location

Cardiff

Print_ISBN

978-1-4244-2709-3

Electronic_ISBN

978-1-4244-2711-6

Type

conf

DOI

10.1109/SSP.2009.5278646

Filename

5278646