Comparing Alternative Methods for Inference in Multiply Sectioned Bayesian Networks

May 2002

Proceedings of the Fifteenth International Florida Artificial Intelligence Research Society Conference (FLAIRS 2002)

Volume

Proceedings of the Fifteenth International Florida Artificial Intelligence Research Society Conference (FLAIRS 2002)

All Papers

Multiply sectioned Bayesian networks (MSBNs) provide one framework for agents to estimate the state of a domain. Existing methods for multi-agent inference in MSBNs are based on linked junction forests (LJFs). The methods are extensions of message passing in junction trees for inference in single-agent Bayesian networks (BNs). We consider extending other inference methods in single-agent BNs to multi-agent inference in MSBNs. In particular, we consider distributed versions of loop cutset conditioning and forward sampling. They are compared with the LJF method in terms of off-line compilation, inter-agent messages during communication, consistent local inference, and preservation of agent privacy.