liu.seSearch for publications in DiVA
Change search
ReferencesLink to record
Permanent link

Direct link
Identification of Gaussian process state-space models with particle stochastic approximation EM
Department of Engineering, University of Cambridge, UK.
Linköping University, Department of Electrical Engineering, Automatic Control. Linköping University, The Institute of Technology.
Deptartment of Information Technology, Uppsala University, Sweden.
Department of Engineering, University of Cambridge, UK.
2014 (English)In: Proceedings of the 19th IFAC World Congress, 2014Conference paper (Refereed)
Abstract [en]

Gaussian process state-space models (GP-SSMs) are a very exible family of models of nonlinear dynamical systems. They comprise a Bayesian nonparametric representation of the dynamics of the system and additional (hyper-)parameters governing the properties of this nonparametric representation. The Bayesian formalism enables systematic reasoning about the uncertainty in the system dynamics. We present an approach to maximum likelihood identification of the parameters in GP-SSMs, while retaining the full nonparametric description of the dynamics. The method is based on a stochastic approximation version of the EM algorithm that employs recent developments in particle Markov chain Monte Carlo for efficient identification.

Place, publisher, year, edition, pages
National Category
Control Engineering
URN: urn:nbn:se:liu:diva-110058OAI: diva2:742563
19th IFAC World Congress, Cape Town, South Africa, August 24-29, 2014.
Available from: 2014-09-02 Created: 2014-09-02 Last updated: 2014-12-17

Open Access in DiVA

fulltext(719 kB)88 downloads
File information
File name FULLTEXT01.pdfFile size 719 kBChecksum SHA-512
Type fulltextMimetype application/pdf

Search in DiVA

By author/editor
Lindsten, Fredrik
By organisation
Automatic ControlThe Institute of Technology
Control Engineering

Search outside of DiVA

GoogleGoogle Scholar
Total: 88 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 184 hits
ReferencesLink to record
Permanent link

Direct link