Relevance Singular Vector Machine for low rank matrix sensing
2014 (English)In: Signal Processing and Communications (SPCOM), 2014 International Conference on, IEEE conference proceedings, 2014, 1-5 p.Conference paper (Refereed)
In this paper we develop a new Bayesian inference method for lowrank matrix reconstruction. We call the new method the RelevanceSingular Vector Machine (RSVM) where appropriate priors are definedon the singular vectors of the underlying matrix to promotelow rank. To accelerate computations, a numerically efficient approximationis developed. The proposed algorithms are applied tomatrix completion and matrix reconstruction problems and their performanceis studied numerically.
Place, publisher, year, edition, pages
IEEE conference proceedings, 2014. 1-5 p.
IdentifiersURN: urn:nbn:se:kth:diva-147601DOI: 10.1109/SPCOM.2014.6983925ISI: 000364936400015ScopusID: 2-s2.0-84920742632ISBN: 978-1-4799-4666-2OAI: oai:DiVA.org:kth-147601DiVA: diva2:730904
International Conference on Signal Processing and Communications (SPCOM), 22-25 July 2014,Bangalore, India
QC 201504132014-06-302014-06-302015-12-18Bibliographically approved