File Download
  Links for fulltext
     (May Require Subscription)
  • Find via Find It@HKUL
Supplementary

Conference Paper: Determining the convergence of variance in Gaussian belief propagation via semi-definite programming

TitleDetermining the convergence of variance in Gaussian belief propagation via semi-definite programming
Authors
Issue Date2014
PublisherIEEE. The Journal's web site is located at http://ieeexplore.ieee.org/xpl/conhome.jsp?punumber=1000369
Citation
The 2014 IEEE International Symposium on Information Theory (ISIT), Honolulu, HI., 29 June-4 July 2014. In IEEE International Symposium on Information Theory Proceedings, 2014, p. 2614-2618 How to Cite?
AbstractIn order to compute the marginal distribution from a high dimensional distribution with loopy Gaussian belief propagation (BP), it is important to determine whether Gaussian BP would converge. In general, the convergence condition for Gaussian BP variance and mean are not necessarily the same, and this paper focuses on the convergence condition of Gaussian BP variance. In particular, by describing the message-passing process of Gaussian BP as a set of updating functions, the necessary and sufficient convergence condition of Gaussian BP variance is derived, with the converged variance proved to be independent of the initialization as long as it is greater or equal to zero. It is further proved that the convergence condition can be verified efficiently by solving a semi-definite programming (SDP) optimization problem. Numerical examples are presented to corroborate the established theories.
Persistent Identifierhttp://hdl.handle.net/10722/199396
ISBN
ISSN

 

DC FieldValueLanguage
dc.contributor.authorSu, Qen_US
dc.contributor.authorWu, YCen_US
dc.date.accessioned2014-07-22T01:15:41Z-
dc.date.available2014-07-22T01:15:41Z-
dc.date.issued2014en_US
dc.identifier.citationThe 2014 IEEE International Symposium on Information Theory (ISIT), Honolulu, HI., 29 June-4 July 2014. In IEEE International Symposium on Information Theory Proceedings, 2014, p. 2614-2618en_US
dc.identifier.isbn978-1-4799-5186-4-
dc.identifier.issn0271-4655-
dc.identifier.urihttp://hdl.handle.net/10722/199396-
dc.description.abstractIn order to compute the marginal distribution from a high dimensional distribution with loopy Gaussian belief propagation (BP), it is important to determine whether Gaussian BP would converge. In general, the convergence condition for Gaussian BP variance and mean are not necessarily the same, and this paper focuses on the convergence condition of Gaussian BP variance. In particular, by describing the message-passing process of Gaussian BP as a set of updating functions, the necessary and sufficient convergence condition of Gaussian BP variance is derived, with the converged variance proved to be independent of the initialization as long as it is greater or equal to zero. It is further proved that the convergence condition can be verified efficiently by solving a semi-definite programming (SDP) optimization problem. Numerical examples are presented to corroborate the established theories.en_US
dc.languageengen_US
dc.publisherIEEE. The Journal's web site is located at http://ieeexplore.ieee.org/xpl/conhome.jsp?punumber=1000369-
dc.relation.ispartofIEEE International Symposium on Information Theory Proceedingsen_US
dc.rightsIEEE International Symposium on Information Theory Proceedings. Copyright © IEEE.-
dc.rights©2014 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.-
dc.rightsCreative Commons: Attribution 3.0 Hong Kong License-
dc.titleDetermining the convergence of variance in Gaussian belief propagation via semi-definite programmingen_US
dc.typeConference_Paperen_US
dc.identifier.emailWu, YC: ycwu@hkucc.hku.hken_US
dc.identifier.authorityWu, YC=rp00195en_US
dc.description.naturepublished_or_final_version-
dc.identifier.hkuros231481en_US
dc.identifier.spage2614-
dc.identifier.epage2618-
dc.publisher.placeUnited States-
dc.customcontrol.immutablesml 140821-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats