BEEM : Bucket Elimination with External Memory
Kalev Kask, Rina Dechter, Andrew Gelfand
A major limitation of exact inference algo- rithms for probabilistic graphical models is their extensive memory usage, which often puts real-world problems out of their reach. In this paper we show how we can extend in- ference algorithms, particularly Bucket Elim- ination, a special case of cluster (join) tree de- composition, to utilize disk memory. We pro- vide the underlying ideas and show promis- ing empirical results of exactly solving large problems not solvable before.
PDF Link: /papers/10/p268-kask.pdf
AUTHOR = "Kalev Kask
and Rina Dechter and Andrew Gelfand",
TITLE = "BEEM : Bucket Elimination with External Memory",
BOOKTITLE = "Proceedings of the Twenty-Sixth Conference Annual Conference on Uncertainty in Artificial Intelligence (UAI-10)",
PUBLISHER = "AUAI Press",
ADDRESS = "Corvallis, Oregon",
YEAR = "2010",
PAGES = "268--276"