Statistical Parameter Learning for Belief Networks with Fixed Structure

Simple item page
dc.contributor.advisorBaras, John S.en_US
dc.contributor.authorLi, Hongjunen_US
dc.contributor.departmentISRen_US
dc.contributor.departmentCSHCNen_US
dc.date.accessioned2007-05-23T10:08:29Z
dc.date.available2007-05-23T10:08:29Z
dc.date.issued1999en_US
dc.description.abstractIn this report, we address the problem of parameter learning for belief networks with fixed structure based on empirical observations. Both complete and incomplete (data) observations are included. Given complete data, we describe the simple problem of single parameter learning for intuition and then expand to belief networks under appropriate system decomposition. If the observations are incomplete, we first estimate the "missing" observations and treat them as though they are "real" observations, based on which the parameter learning can be executed as in complete data case. We derive a uniform algorithm based on this idea for incomplete data case and present the convergence and optimality properties. Such an algorithm is suitable trivially under complete observations.en_US
dc.format.extent512893 bytes
dc.format.mimetypeapplication/pdf
dc.identifier.urihttp://hdl.handle.net/1903/6089
dc.language.isoen_USen_US
dc.relation.ispartofseriesISR; TR 1999-59en_US
dc.relation.ispartofseriesCSHCN; TR 1999-32en_US
dc.subjectestimationen_US
dc.subjectnetwork managementen_US
dc.subjectknowledge representationen_US
dc.subjectbelief networksen_US
dc.subjectmaximum likelihooden_US
dc.subjectEM algorithmen_US
dc.subjectIntelligent Signal Processing and Communications Systemsen_US
dc.subjectSystems Integration Methodologyen_US
dc.titleStatistical Parameter Learning for Belief Networks with Fixed Structureen_US
dc.typeTechnical Reporten_US

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
TR_99-59.pdf
Size:
500.87 KB
Format:
Adobe Portable Document Format
Download

Collections

Institute for Systems Research Technical Reports