Bayesian Inference for Structured Spike and Slab Priors

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

Sparse signal recovery addresses the problem of solving underdetermined linear inverse problems subject to a sparsity constraint. We propose a novel prior formulation, the structured spike and slab prior, which allows to incorporate a priori knowledge of the sparsity pattern by imposing a spatial Gaussian process on the spike and slab probabilities. Thus, prior information on the structure of the sparsity pattern can be encoded using generic covariance functions. Furthermore, we provide a Bayesian inference scheme for the proposed model based on the expectation propagation framework. Using numerical experiments on synthetic data, we demonstrate the benefits of the model.
Original languageEnglish
Title of host publicationProceedings of the 28th Annual Conference on Advances in Neural Information Processing Systems 27 (NIPS 2014)
PublisherNeural Information Processing Systems Foundation
Publication date2014
Pages1745-1753
Publication statusPublished - 2014
Event28th Annual Conference on Neural Information Processing Systems (NIPS 2014) - Montréal, Canada
Duration: 8 Dec 201413 Dec 2014
Conference number: 28
https://nips.cc/Conferences/2014

Conference

Conference28th Annual Conference on Neural Information Processing Systems (NIPS 2014)
Number28
CountryCanada
CityMontréal
Period08/12/201413/12/2014
Internet address

Fingerprint Dive into the research topics of 'Bayesian Inference for Structured Spike and Slab Priors'. Together they form a unique fingerprint.

Cite this