Files in this item
Faster inference from state space models via GPU computing
Item metadata
dc.contributor.author | Fagard-Jenkin, Calliste | |
dc.contributor.author | Thomas, Len | |
dc.date.accessioned | 2024-02-02T12:30:04Z | |
dc.date.available | 2024-02-02T12:30:04Z | |
dc.date.issued | 2024-05 | |
dc.identifier | 298572445 | |
dc.identifier | 0ab0880e-65ec-4a36-a5a1-643e2edd5836 | |
dc.identifier | 85184024660 | |
dc.identifier.citation | Fagard-Jenkin , C & Thomas , L 2024 , ' Faster inference from state space models via GPU computing ' , Ecological Informatics , vol. 80 , 102486 . https://doi.org/10.1016/j.ecoinf.2024.102486 | en |
dc.identifier.issn | 1574-9541 | |
dc.identifier.other | RIS: urn:811050C074FBCC2EB042BD601BD5BEE8 | |
dc.identifier.other | ORCID: /0000-0001-6786-4936/work/152318027 | |
dc.identifier.other | ORCID: /0000-0002-7436-067X/work/152318282 | |
dc.identifier.uri | https://hdl.handle.net/10023/29141 | |
dc.description | Funding: C.F.-J. is funded via a doctoral scholarship from the University of St Andrews, School of Mathematics and Statistics. | en |
dc.description.abstract | Inexpensive Graphics Processing Units (GPUs) offer the potential to greatly speed up computation by employing their massively parallel architecture to perform arithmetic operations more efficiently. Population dynamics models are important tools in ecology and conservation. Modern Bayesian approaches allow biologically realistic models to be constructed and fitted to multiple data sources in an integrated modelling framework based on a class of statistical models called state space models. However, model fitting is often slow, requiring hours to weeks of computation. We demonstrate the benefits of GPU computing using a model for the population dynamics of British grey seals, fitted with a particle Markov chain Monte Carlo algorithm. Speed-ups of two orders of magnitude were obtained for estimations of the log-likelihood, compared to a traditional ‘CPU-only’ implementation, allowing for an accurate method of inference to be used where this was previously too computationally expensive to be viable. GPU computing has enormous potential, but one barrier to further adoption is a steep learning curve, due to GPUs' unique hardware architecture. We provide a detailed description of hardware and software setup, and our case study provides a template for other similar applications. We also provide a detailed tutorial-style description of GPU hardware architectures, and examples of important GPU-specific programming practices. | |
dc.format.extent | 25 | |
dc.format.extent | 9434764 | |
dc.language.iso | eng | |
dc.relation.ispartof | Ecological Informatics | en |
dc.subject | Bayesian inference | en |
dc.subject | CUDA | en |
dc.subject | GPU | en |
dc.subject | Grey seal | en |
dc.subject | Particle filter | en |
dc.subject | Parallel processing | en |
dc.subject | Particle Markov chain Monte Carlo | en |
dc.subject | Population dynamics model | en |
dc.subject | QA Mathematics | en |
dc.subject | DAS | en |
dc.subject.lcc | QA | en |
dc.title | Faster inference from state space models via GPU computing | en |
dc.type | Journal article | en |
dc.contributor.institution | University of St Andrews. Statistics | en |
dc.contributor.institution | University of St Andrews. Centre for Research into Ecological & Environmental Modelling | en |
dc.contributor.institution | University of St Andrews. Marine Alliance for Science & Technology Scotland | en |
dc.identifier.doi | 10.1016/j.ecoinf.2024.102486 | |
dc.description.status | Peer reviewed | en |
This item appears in the following Collection(s)
Items in the St Andrews Research Repository are protected by copyright, with all rights reserved, unless otherwise indicated.