The Library
Spatio-temporal variational Gaussian processes
Tools
Hamelijnck, Oliver, Wilkinson, William J., Loppi, Niki A., Solin, Arno and Damoulas, Theodoros (2021) Spatio-temporal variational Gaussian processes. In: Thirty-fifth Conference on Neural Information Processing Systems (NeurIPS 2021), Virtual, 6-14 Dec 2021. Published in: Advances in Neural Information Processing Systems 34 (NeurIPS 2021), 34 pp. 23621-23633.
|
PDF
WRAP-spatio-temporal-variational-Gaussian-processes-Damoulas-v2-2021.pdf - Accepted Version - Requires a PDF viewer. Download (6Mb) | Preview |
|
PDF
WRAP-spatio-temporal-variational-Gaussian-processes-Damoulas-2021.pdf - Accepted Version Embargoed item. Restricted access to Repository staff only - Requires a PDF viewer. Download (2066Kb) |
Official URL: https://proceedings.neurips.cc/paper/2021/file/c6b...
Abstract
We introduce a scalable approach to Gaussian process inference that combines spatio-temporal filtering with natural gradient variational inference, resulting in a non-conjugate GP method for multivariate data that scales linearly with respect to time. Our natural gradient approach enables application of parallel filtering and smoothing, further reducing the temporal span complexity to be logarithmic in the number of time steps. We derive a sparse approximation that constructs a state-space model over a reduced set of spatial inducing points, and show that for separable Markov kernels the full and sparse cases exactly recover the standard variational GP, whilst exhibiting favourable computational properties. To further improve the spatial scaling we propose a mean-field assumption of independence between spatial locations which, when coupled with sparsity and parallelisation, leads to an efficient and accurate method for large spatio-temporal problems.
Item Type: | Conference Item (Paper) | |||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Subjects: | Q Science > QA Mathematics | |||||||||||||||
Divisions: | Faculty of Science, Engineering and Medicine > Science > Computer Science | |||||||||||||||
Library of Congress Subject Headings (LCSH): | Gaussian processes, Machine learning -- Mathematical models, Gaussian processes -- Data processing | |||||||||||||||
Journal or Publication Title: | Advances in Neural Information Processing Systems 34 (NeurIPS 2021) | |||||||||||||||
Publisher: | Curran Associates, Inc. | |||||||||||||||
Official Date: | 2021 | |||||||||||||||
Dates: |
|
|||||||||||||||
Volume: | 34 | |||||||||||||||
Page Range: | pp. 23621-23633 | |||||||||||||||
Status: | Peer Reviewed | |||||||||||||||
Publication Status: | Published | |||||||||||||||
Access rights to Published version: | Open Access (Creative Commons) | |||||||||||||||
Date of first compliant deposit: | 20 October 2021 | |||||||||||||||
Date of first compliant Open Access: | 20 October 2021 | |||||||||||||||
RIOXX Funder/Project Grant: |
|
|||||||||||||||
Conference Paper Type: | Paper | |||||||||||||||
Title of Event: | Thirty-fifth Conference on Neural Information Processing Systems (NeurIPS 2021) | |||||||||||||||
Type of Event: | Conference | |||||||||||||||
Location of Event: | Virtual | |||||||||||||||
Date(s) of Event: | 6-14 Dec 2021 | |||||||||||||||
Related URLs: |
Request changes or add full text files to a record
Repository staff actions (login required)
View Item |
Downloads
Downloads per month over past year