Stratified exponential families: Graphical models and model selection
Geiger, Dan ; Heckerman, David ; King, Henry ; Meek, Christopher
Ann. Statist., Tome 29 (2001) no. 2, p. 505-529 / Harvested from Project Euclid
We describe a hierarchy of exponential families which is useful for distinguishing types of graphical models. Undirected graphical models with no hidden variables are linear exponential families (LEFs). Directed acyclic graphical (DAG) models and chain graphs with no hidden variables, includ­ ing DAG models with several families of local distributions, are curved exponential families (CEFs). Graphical models with hidden variables are what we term stratified exponential families (SEFs). A SEF is a finite union of CEFs of various dimensions satisfying some regularity conditions. We also show that this hierarchy of exponential families is noncollapsing with respect to graphical models by providing a graphical model which is a CEF but not a LEF and a graphical model that is a SEF but not a CEF. Finally, we show how to compute the dimension of a stratified exponential family. These results are discussed in the context of model selection of graphical models.
Publié le : 2001-04-14
Classification:  Bayesian networks,  graphical models,  hidden variables,  curved exponential families,  stratified exponential families,  semialgebraic sets,  model selection,  60E05,  62H05
@article{1009210550,
     author = {Geiger, Dan and Heckerman, David and King, Henry and Meek, Christopher},
     title = {Stratified exponential families: Graphical models and model
			 selection},
     journal = {Ann. Statist.},
     volume = {29},
     number = {2},
     year = {2001},
     pages = { 505-529},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1009210550}
}
Geiger, Dan; Heckerman, David; King, Henry; Meek, Christopher. Stratified exponential families: Graphical models and model
			 selection. Ann. Statist., Tome 29 (2001) no. 2, pp.  505-529. http://gdmltest.u-ga.fr/item/1009210550/