English
 
Help Privacy Policy Disclaimer
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT
  Expectation Propagation, Experimental Design for the Sparse Linear Model

Seeger, M. (2008). Expectation Propagation, Experimental Design for the Sparse Linear Model. Talk presented at University of Cambridge: Engineering Department. Cambridge, UK. 2008-02-20.

Item is

Files

show Files

Locators

show
hide
Description:
-

Creators

show
hide
 Creators:
Seeger, M1, 2, Author              
Affiliations:
1Department Empirical Inference, Max Planck Institute for Biological Cybernetics, Max Planck Society, ou_1497795              
2Max Planck Institute for Biological Cybernetics, Max Planck Society, Spemannstrasse 38, 72076 Tübingen, DE, ou_1497794              

Content

show
hide
Free keywords: -
 Abstract: Expectation propagation (EP) is a novel variational method for approximate Bayesian inference, which has given promising results in terms of computational efficiency and accuracy in several machine learning applications. It can readily be applied to inference in linear models with non-Gaussian priors, generalised linear models, or nonparametric Gaussian process models, among others. I will give an introduction to this framework. Important aspects of EP are not well-understood theoretically. I will highlight some open problems. I will then show how to address sequential experimental design for a linear model with non-Gaussian sparsity priors, giving some results in two different machine learning applications. These results indicate that experimental design for these models may have significantly different properties than for linear-Gaussian models, where Bayesian inference is analytically tractable and experimental design seems best understood.

Details

show
hide
Language(s):
 Dates: 2008-02
 Publication Status: Published online
 Pages: -
 Publishing info: -
 Table of Contents: -
 Rev. Type: -
 Identifiers: -
 Degree: -

Event

show
hide
Title: University of Cambridge: Engineering Department
Place of Event: Cambridge, UK
Start-/End Date: 2008-02-20
Invited: Yes

Legal Case

show

Project information

show

Source

show