Abstract
It is shown that for every undirected graph G over a finite set N and for every nonempty T ⊂ N there exists an undirected graph G T over T, called the marginal graph of G for T, such that the class of marginal distributions for T of (discrete) G-Markovian distributions coincides with the class of G T-Markovian distributions. An example shows that this is not true within the framework of strictly positive probability distributions. However, an analogous positive result holds for hypergraphs and classes of strictly positive factorizable distributions.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Asmussen, S. and Edwards, D. (1983) Collapsibility and response variables in contingency tables, Biometrika, 70, pp. 567–578.
Frydenberg, M. (1990) Marginalization and collapsibility in graphical interaction models, The Annals Statistics, 18, pp. 790–805.
Lauritzen, S.L., Dawid, A.P., Larsen B.N. and Leimer, H.-G. (1990) Independence properties of directed Markov fields, Networks, 20, pp. 491–505.
Malvestuto, F.M. (1996) Testing implication of hierarchical log-linear models for probability distributions, Statistics and Computing, 6, pp. 169–176.
Pearl, J. (1988) Probabilistic Reasoning in Intelligent Systems. Morgan Kaufmann, San Mateo.
Speed, T.P. (1979) A note on nearest-neighbour Gibbs and Markov probabilities, Sakhya A, 41, pp. 184–197.
Studený, M. (1994/1995) Description of structures of stochastic conditional independence by means of faces and imsets (a series of 3 papers), International Journal of General Systems, 23, pp. 123–137, 201–219, 323–341.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1997 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Studený, M. (1997). On Marginalization, Collapsibility and Precollapsibility. In: Beneš, V., Štěpán, J. (eds) Distributions with given Marginals and Moment Problems. Springer, Dordrecht. https://doi.org/10.1007/978-94-011-5532-8_22
Download citation
DOI: https://doi.org/10.1007/978-94-011-5532-8_22
Publisher Name: Springer, Dordrecht
Print ISBN: 978-94-010-6329-6
Online ISBN: 978-94-011-5532-8
eBook Packages: Springer Book Archive