Population Monte Carlo

Cappé, Olivier; Guillin, Arnaud; Marin, Jean-Michel; Robert, Christian P.

Cappé, Olivier; Guillin, Arnaud; Marin, Jean-Michel; Robert, Christian P. (2004), Population Monte Carlo, Journal of Computational and Graphical Statistics, 13, 4, p. 907-929. http://dx.doi.org/10.1198/106186004X12803

View/Open

2002-34.ps (3.298Mb)

popMC.PDF (1.150Mb)

Type

Article accepté pour publication ou publié

Date

2004

Journal name

Journal of Computational and Graphical Statistics

Volume

Number

Publisher

American Statistical Association

Pages

907-929

Abstract (EN)

Importance sampling methods can be iterated like MCMC algorithms, while being more robust against dependence and starting values. The population Monte Carlo principle consists of iterated generations of importance samples, with importance functions depending on the previously generated importance samples. The advantage over MCMC algorithms is that the scheme is unbiased at any iteration and can thus be stopped at any time, while iterations improve the performances of the importance function, thus leading to an adaptive importance sampling. We illustrate this method on a mixture example with multiscale importance functions. A second example reanalyzes the ion channel model using an importance sampling scheme based on a hidden Markov representation, and compares population Monte Carlo with a corresponding MCMC algorithm.

Subjects / Keywords

Multiple scales; Ion channel model; importance sampling; hidden semi-Markov model; adaptive algorithms

JEL

C11 - Bayesian Analysis: General