Multilevel Monte Carlo in Approximate Bayesian Computation
Methodology
2017-02-14 v1
Abstract
In the following article we consider approximate Bayesian computation (ABC) inference. We introduce a method for numerically approximating ABC posteriors using the multilevel Monte Carlo (MLMC). A sequential Monte Carlo version of the approach is developed and it is shown under some assumptions that for a given level of mean square error, this method for ABC has a lower cost than i.i.d. sampling from the most accurate ABC approximation. Several numerical examples are given.
Cite
@article{arxiv.1702.03628,
title = {Multilevel Monte Carlo in Approximate Bayesian Computation},
author = {Ajay Jasra and Seongil Jo and David Nott and Christine Shoemaker and Raul Tempone},
journal= {arXiv preprint arXiv:1702.03628},
year = {2017}
}
Comments
This paper was submitted in Dec 2016