Efficient state-space inference of periodic latent force models
- Submitting institution
-
University of Oxford
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 1946
- Type
- D - Journal article
- DOI
-
10.5555/2627435.2670325
- Title of journal
- Journal of Machine Learning Research
- Article number
- -
- First page
- 2337
- Volume
- 15
- Issue
- -
- ISSN
- 1532-4435
- Open access status
- Out of scope for open access requirements
- Month of publication
- July
- Year of publication
- 2014
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
4
- Research group(s)
-
-
- Citation count
- 6
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This paper develops a computationally efficient representation of latent force models that allows their use in previously prohibitively computationally expensive settings, such as those where the forces are periodic. The approach is applied to modelling the thermal properties of buildings subject to unknown periodic heating effects, such as solar heating, and is shown to be more effective in predicting day-ahead temperatures within a home than other state-of-the-art methods. The work here informed the subsequent thermal modelling within a spin-out company, called Joulo, co-founded by Rogers, that used thermal modelling to provide personalised home heating advice to households.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -