Multilevel Monte Carlo in Approximate Bayesian Computation

Thumbnail
Name:
1702.03628v1.pdf
Size:
391.1Kb
Format:
PDF
Description:
Preprint
Download
Type
Preprint
Authors
Jasra, Ajay
Jo, Seongil
Nott, David
Shoemaker, Christine
Tempone, Raul cc
KAUST Department
Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
Applied Mathematics and Computational Science Program
Center for Uncertainty Quantification in Computational Science & Engineering, King Abdullah University of Science and Technology, Thuwal, 23955-6900, KSA.
Date
2017-02-13

Metadata
Show full item record
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.
Publisher
arXiv
arXiv
arXiv:1702.03628
Handle URI
http://hdl.handle.net/10754/626522
Additional Links
http://arxiv.org/abs/1702.03628v1
http://arxiv.org/pdf/1702.03628v1
Collections
Other/General Submission; Applied Mathematics and Computational Science Program; Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division