Asymptotic Model Selection for Directed Networks with Hidden Variables
Dan Geiger, David Heckerman, Christopher Meek
We extend the Bayesian Information Criterion (BIC), an asymptotic approximation for the marginal likelihood, to Bayesian networks with hidden variables. This approximation can be used to select models given large samples of data. The standard BIC as well as our extension punishes the complexity of a model according to the dimension of its parameters. We argue that the dimension of a Bayesian network with hidden variables is the rank of the Jacobian matrix of the transformation between the parameters of the network and the parameters of the observable variables. We compute the dimensions of several networks including the naive Bayes model with a hidden root node.
Keywords: Bayesian model selection, marginal likelihood, asymptotic approximations.
PS Link: ftp://ftp.research.microsoft.com/pub/Tech-Reports/Winter95-96/TR-96-07.PS
PDF Link: /papers/96/p283-geiger.pdf
AUTHOR = "Dan Geiger
and David Heckerman and Christopher Meek",
TITLE = "Asymptotic Model Selection for Directed Networks with Hidden Variables",
BOOKTITLE = "Proceedings of the Twelfth Conference Annual Conference on Uncertainty in Artificial Intelligence (UAI-96)",
PUBLISHER = "Morgan Kaufmann",
ADDRESS = "San Francisco, CA",
YEAR = "1996",
PAGES = "283--290"