Online Gaussian Process regression with non-Gaussian likelihood

David Seiferth, G. Chowdhary, M. Muhlegg, F. Holzapfel

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We present a new algorithm for GP regression over data with non-Gaussian likelihood that does not require costly MCMC sampling, or variational Bayes optimization. In our method, which we term Meta-GP, we model the likelihood by another Gaussian Process point-wise in time. This approach allows for the calculation of the posterior predictive mean and variance in an analytical way pointwise in time, leading to an online inference algorithm. As a result, our method can work with streaming data, is analytically tractable, computationally efficient while being as accurate or better than Expectation Propagation, Laplace Approximation, and MCMC inference methods for non-Gaussian likelihood data.

Original languageEnglish (US)
Title of host publication2017 American Control Conference, ACC 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages3134-3140
Number of pages7
ISBN (Electronic)9781509059928
DOIs
StatePublished - Jun 29 2017
Event2017 American Control Conference, ACC 2017 - Seattle, United States
Duration: May 24 2017May 26 2017

Publication series

NameProceedings of the American Control Conference
ISSN (Print)0743-1619

Other

Other2017 American Control Conference, ACC 2017
Country/TerritoryUnited States
CitySeattle
Period5/24/175/26/17

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Online Gaussian Process regression with non-Gaussian likelihood'. Together they form a unique fingerprint.

Cite this