×

zbMATH — the first resource for mathematics

Local score computation in learning belief networks. (English) Zbl 0984.68514
Stroulia, Eleni (ed.) et al., Advances in artificial intelligence. 14th biennial conference of the Canadian society for computational studies of intelligence, AI 2001, Ottawa, Canada, June 7-9, 2001. Proceedings. Berlin: Springer. Lect. Notes Comput. Sci. 2056, 152-161 (2001).
Summary: We propose an improved scoring metrics for learning belief networks driven by issues arising from learning in pseudo-independent domains. We identify a small subset of variables called a crux, which is sufficient to compute the incremental improvement of alternative belief network structures. We prove formally that such local computation, while improving efficiency, does not introduce any error to the evaluation of alternative structures.
For the entire collection see [Zbl 0967.00072].
MSC:
68T05 Learning and adaptive systems in artificial intelligence
68T37 Reasoning under uncertainty in the context of artificial intelligence
68P20 Information storage and retrieval of data
Software:
WEBWEAVR-III
PDF BibTeX XML Cite
Full Text: Link