Generalized graphical models for discrete data

Traditional graphical models are extended by allowing that the presence or absence of a connection between two nodes depends on the values of the remaining variables. We first compare the extended model to the classical log-linear model. After discussing the induced consistency problem we illustrate the corresponding estimation problem by way of an example.

MoreLess

Year of publication:	1998
Authors:	Teugels, Jozef L. ; Van Horebeek, Johan
Published in:	Statistics & Probability Letters. - Elsevier, ISSN 0167-7152. - Vol. 38.1998, 1, p. 41-47
Publisher:	Elsevier
Keywords:	Graphical model Conditional independency Markov random field Log-linear model

More details

Type of publication:	Article
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10005319538