Combined modeling of sparse and dense noise improves Bayesian RVM
2014 (English)In: European Signal Processing Conference, 2014, 1841-1845 p.Conference paper (Refereed)
Using a Bayesian approach, we consider the problem of recovering sparse signals under additive sparse and dense noise. Typically, sparse noise models outliers, impulse bursts or data loss. To handle sparse noise, existing methods simultaneously estimate sparse noise and sparse signal of interest. For estimating the sparse signal, without estimating the sparse noise, we construct a Relevance Vector Machine (RVM). In the RVM, sparse noise and ever present dense noise are treated through a combined noise model. Through simulations, we show the efficiency of new RVM for three applications: kernel regression, housing price prediction and compressed sensing.
Place, publisher, year, edition, pages
2014. 1841-1845 p.
Bayesian learning, Compressed sensing, Relevance vector machine, Robust regression, Bayesian networks, Housing, Signal processing, Signal reconstruction, Bayesian approaches, Combined model, Combined noise models, Kernel regression, Robust regressions, Sparse signals
IdentifiersURN: urn:nbn:se:kth:diva-167957ScopusID: 2-s2.0-84911921506ISBN: 9780992862619OAI: oai:DiVA.org:kth-167957DiVA: diva2:817113
22nd European Signal Processing Conference, EUSIPCO 2014, 1 September 2014 through 5 September 2014
QC 201506042015-06-042015-05-222016-04-27Bibliographically approved