Structured priors for structure learning

V. K. Mansinghka, C. Kemp, J. B. Tenenbaum, Thomas L. Griffiths

Research output: Chapter in Book/Report/Conference proceedingConference contribution

28 Scopus citations

Abstract

Traditional approaches to Bayes net structure learning typically assume little regularity in graph structure other than sparseness. However, in many cases, we expect more systematicity: variables in real-world systems often group into classes that predict the kinds of probabilistic dependencies they participate in. Here we capture this form of prior knowledge in a hierarchical Bayesian framework, and exploit it to enable structure learning and type discovery from small datasets. Specifically, we present a nonparametric generative model for directed acyclic graphs as a prior for Bayes net structure learning. Our model assumes that variables come in one or more classes and that the prior probability of an edge existing between two variables is a function only of their classes. We derive an MCMC algorithm for simultaneous inference of the number of classes, the class assignments of variables, and the Bayes net structure over variables. For several realistic, sparse datasets, we show that the bias towards systematicity of connections provided by our model can yield more accurate learned networks than the traditional approach of using a uniform prior, and that the classes found by our model are appropriate.

Original languageEnglish (US)
Title of host publicationProceedings of the 22nd Conference on Uncertainty in Artificial Intelligence, UAI 2006
Pages324-331
Number of pages8
StatePublished - Dec 1 2006
Externally publishedYes
Event22nd Conference on Uncertainty in Artificial Intelligence, UAI 2006 - Cambridge, MA, United States
Duration: Jul 13 2006Jul 16 2006

Publication series

NameProceedings of the 22nd Conference on Uncertainty in Artificial Intelligence, UAI 2006

Other

Other22nd Conference on Uncertainty in Artificial Intelligence, UAI 2006
CountryUnited States
CityCambridge, MA
Period7/13/067/16/06

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence

Cite this

Mansinghka, V. K., Kemp, C., Tenenbaum, J. B., & Griffiths, T. L. (2006). Structured priors for structure learning. In Proceedings of the 22nd Conference on Uncertainty in Artificial Intelligence, UAI 2006 (pp. 324-331). (Proceedings of the 22nd Conference on Uncertainty in Artificial Intelligence, UAI 2006).