Sequential MCMC for Bayesian model selection

Author

Andrieu, Christophe ; De Freitas, Nando ; Doucet, Arnaud

Author_Institution

Dept. of Eng., Cambridge Univ., UK

fYear

1999

fDate

1999

Firstpage

130

Lastpage

134

Abstract

In this paper, we address the problem of sequential Bayesian model selection. This problem does not usually admit any closed-form analytical solution. We propose here an original sequential simulation-based method to solve the associated Bayesian computational problems. This method combines sequential importance sampling, a resampling procedure and reversible jump MCMC (Markov chain Monte Carlo) moves. We describe a generic algorithm and then apply it to the problem of sequential Bayesian model order estimation of autoregressive (AR) time series observed in additive noise

Keywords

Bayes methods; Markov processes; Monte Carlo methods; autoregressive processes; importance sampling; noise; parameter estimation; signal sampling; time series; AR time series; Bayesian computational problems; Markov chain Monte Carlo methods; autoregressive time series; generic algorithm; order estimation; resampling procedure; reversible jump MCMC moves; sequential Bayesian model selection; sequential MCMC; sequential importance sampling; sequential simulation-based method; signal model; signal sampling; Bayesian methods; Computational modeling; Data analysis; Electronic switching systems; Gaussian noise; Identity-based encryption; Monte Carlo methods; Nominations and elections; State estimation; Stochastic processes;

fLanguage

English

Publisher

ieee

Conference_Titel

Higher-Order Statistics, 1999. Proceedings of the IEEE Signal Processing Workshop on

Conference_Location

Caesarea

Print_ISBN

0-7695-0140-0

Type

conf

DOI

10.1109/HOST.1999.778709

Filename

778709