State-Space Inference and Learning with Gaussian Processes

Ryan Turner, Marc Deisenroth, Carl Rasmussen
; Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, JMLR Workshop and Conference Proceedings 9:868-875, 2010.

Abstract

State-space inference and learning with Gaussian processes (GPs) is an unsolved problem. We propose a new, general methodology for inference and learning in nonlinear state-space models that are described probabilistically by non-parametric GP models. We apply the expectation maximization algorithm to iterate between inference in the latent state-space and learning the parameters of the underlying GP dynamics model.

Cite this Paper


BibTeX
@InProceedings{pmlr-v9-turner10a, title = {State-Space Inference and Learning with Gaussian Processes}, author = {Ryan Turner and Marc Deisenroth and Carl Rasmussen}, booktitle = {Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics}, pages = {868--875}, year = {2010}, editor = {Yee Whye Teh and Mike Titterington}, volume = {9}, series = {Proceedings of Machine Learning Research}, address = {Chia Laguna Resort, Sardinia, Italy}, month = {13--15 May}, publisher = {JMLR Workshop and Conference Proceedings}, pdf = {http://proceedings.mlr.press/v9/turner10a/turner10a.pdf}, url = {http://proceedings.mlr.press/v9/turner10a.html}, abstract = {State-space inference and learning with Gaussian processes (GPs) is an unsolved problem. We propose a new, general methodology for inference and learning in nonlinear state-space models that are described probabilistically by non-parametric GP models. We apply the expectation maximization algorithm to iterate between inference in the latent state-space and learning the parameters of the underlying GP dynamics model.} }
Endnote
%0 Conference Paper %T State-Space Inference and Learning with Gaussian Processes %A Ryan Turner %A Marc Deisenroth %A Carl Rasmussen %B Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2010 %E Yee Whye Teh %E Mike Titterington %F pmlr-v9-turner10a %I PMLR %J Proceedings of Machine Learning Research %P 868--875 %U http://proceedings.mlr.press %V 9 %W PMLR %X State-space inference and learning with Gaussian processes (GPs) is an unsolved problem. We propose a new, general methodology for inference and learning in nonlinear state-space models that are described probabilistically by non-parametric GP models. We apply the expectation maximization algorithm to iterate between inference in the latent state-space and learning the parameters of the underlying GP dynamics model.
RIS
TY - CPAPER TI - State-Space Inference and Learning with Gaussian Processes AU - Ryan Turner AU - Marc Deisenroth AU - Carl Rasmussen BT - Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics PY - 2010/03/31 DA - 2010/03/31 ED - Yee Whye Teh ED - Mike Titterington ID - pmlr-v9-turner10a PB - PMLR SP - 868 DP - PMLR EP - 875 L1 - http://proceedings.mlr.press/v9/turner10a/turner10a.pdf UR - http://proceedings.mlr.press/v9/turner10a.html AB - State-space inference and learning with Gaussian processes (GPs) is an unsolved problem. We propose a new, general methodology for inference and learning in nonlinear state-space models that are described probabilistically by non-parametric GP models. We apply the expectation maximization algorithm to iterate between inference in the latent state-space and learning the parameters of the underlying GP dynamics model. ER -
APA
Turner, R., Deisenroth, M. & Rasmussen, C.. (2010). State-Space Inference and Learning with Gaussian Processes. Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, in PMLR 9:868-875

Related Material