Fast Convergent Algorithms for Expectation Propagation Approximate Bayesian Inference

2011

Conference Paper

ei


We propose a novel algorithm to solve the expectation propagation relaxation of Bayesian inference for continuous-variable graphical models. In contrast to most previous algorithms, our method is provably convergent. By marrying convergent EP ideas from (Opper&Winther, 2005) with covariance decoupling techniques (Wipf&Nagarajan, 2008; Nickisch&Seeger, 2009), it runs at least an order of magnitude faster than the most common EP solver.

Author(s): Seeger, M. and Nickisch, H.
Book Title: JMLR Workshop and Conference Proceedings Volume 15: AISTATS 2011
Pages: 652-660
Year: 2011
Month: April
Day: 0
Editors: Gordon, G. , D. Dunson, M. Dudík
Publisher: MIT Press

Department(s): Empirical Inference
Research Project(s): Probabilistic Inference
Bibtex Type: Conference Paper (inproceedings)

Address: Cambridge, MA, USA
Digital: 0
Event Name: 14th International Conference on Artificial Intelligence and Statistics
Event Place: Fort Lauderdale, FL, USA

Links: PDF
Web

BibTex

@inproceedings{SeegerN2011,
  title = {Fast Convergent Algorithms for Expectation Propagation Approximate Bayesian Inference },
  author = {Seeger, M. and Nickisch, H.},
  booktitle = {JMLR Workshop and Conference Proceedings Volume 15: AISTATS 2011},
  pages = {652-660},
  editors = {Gordon, G. , D. Dunson, M. Dudík },
  publisher = {MIT Press},
  address = {Cambridge, MA, USA},
  month = apr,
  year = {2011},
  month_numeric = {4}
}