Efficient Search-Based Inference for Noisy-OR Belief Networks: TopEpsilon
Kurt Huang, Max Henrion
Inference algorithms for arbitrary belief networks are impractical for large, complex belief networks. Inference algorithms for specialized classes of belief networks have been shown to be more efficient. In this paper, we present a search-based algorithm for approximate inference on arbitrary, noisy-OR belief networks, generalizing earlier work on search-based inference for two-level, noisy-OR belief networks. Initial experimental results appear promising.
Keywords: Noisy-OR belief networks, search-based inference, CPCS.
PS Link: http://www-smi.stanford.edu/people/khuang/huangk-uai96.ps
PDF Link: /papers/96/p325-huang.pdf
AUTHOR = "Kurt Huang
and Max Henrion",
TITLE = "Efficient Search-Based Inference for Noisy-OR Belief Networks: TopEpsilon",
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 = "325--331"