The Library
SigGPDE : scaling sparse Gaussian processes on sequential data
Tools
Lemercier, Maud, Salvi, Cristopher, Cass, Thomas, Bonilla, Edwin V., Damoulas, Theodoros and Lyons, Terry (2021) SigGPDE : scaling sparse Gaussian processes on sequential data. In: ICML : 2021 Thirty-eighth International Conference on Machine Learning, Virtual, 18-24 Jul 2021. Published in: Proceedings of the 38th International Conference on Machine Learning, 139 pp. 6233-6242. ISSN 2640-3498.
|
PDF
WRAP-SigGPDE-scaling-sparse-Gaussian-processes-sequential-data-2021.pdf - Accepted Version - Requires a PDF viewer. Download (1143Kb) | Preview |
Official URL: https://proceedings.mlr.press/v139/lemercier21a.ht...
Abstract
Making predictions and quantifying their uncertainty when the input data is sequential is a fundamental learning challenge, recently attracting increasing attention. We develop SigGPDE, a new scalable sparse variational inference framework for Gaussian Processes (GPs) on sequential data. Our contribution is twofold. First, we construct inducing variables underpinning the sparse approximation so that the resulting evidence lower bound (ELBO) does not require any matrix inversion. Second, we show that the gradients of the GP signature kernel are solutions of a hyperbolic partial differential equation (PDE). This theoretical insight allows us to build an efficient back-propagation algorithm to optimize the ELBO. We showcase the significant computational gains of SigGPDE compared to existing methods, while achieving state-of-the-art performance for classification tasks on large datasets of up to 1 million multivariate time series.
Item Type: | Conference Item (Paper) | ||||||
---|---|---|---|---|---|---|---|
Subjects: | Q Science > Q Science (General) Q Science > QA Mathematics |
||||||
Divisions: | Faculty of Science, Engineering and Medicine > Science > Computer Science | ||||||
Library of Congress Subject Headings (LCSH): | Gaussian processes -- Data processing, Differential equations, Partial, Machine learning -- Mathematical models | ||||||
Series Name: | Proceedings of Machine Learning Research | ||||||
Journal or Publication Title: | Proceedings of the 38th International Conference on Machine Learning | ||||||
Publisher: | PMLR | ||||||
ISSN: | 2640-3498 | ||||||
Official Date: | 2021 | ||||||
Dates: |
|
||||||
Volume: | 139 | ||||||
Page Range: | pp. 6233-6242 | ||||||
Status: | Peer Reviewed | ||||||
Publication Status: | Published | ||||||
Access rights to Published version: | Restricted or Subscription Access | ||||||
Date of first compliant deposit: | 3 June 2021 | ||||||
Date of first compliant Open Access: | 3 June 2021 | ||||||
Conference Paper Type: | Paper | ||||||
Title of Event: | ICML : 2021 Thirty-eighth International Conference on Machine Learning | ||||||
Type of Event: | Conference | ||||||
Location of Event: | Virtual | ||||||
Date(s) of Event: | 18-24 Jul 2021 | ||||||
Related URLs: | |||||||
Open Access Version: |
Request changes or add full text files to a record
Repository staff actions (login required)
View Item |
Downloads
Downloads per month over past year