The CLP(BN) System
CLP(BN) is a novel approach that aims at
integrating Bayesian Networks with the constraint logic programming
framework. The system was designed by Vítor Santos Costa, David Page, and James Cussens.
CLP(BN) relies on Yap as the underlying Prolog
engine.
This page is currently being built. For now we have available:
- click here for a description of the ideas behind
CLP(BN). A shorter version of this paper was accepted at UAI2003.
- check here for the current
CLP(BN) prototype (Yap must have coroutining
enabled). While CLP(BN) is self-contained, it has a modular design so
that its current constraint solver (Bayes net inference engine) can
be replaced by alternative algorithms for answering Bayes net queries,
such as the variable elimination routine in Kevin Murphy's BN
toolbox.
Please also have look at Kristian Kersting's
Probabilistic-Logical Models Repository for an excellent set of links
on related work.
Vítor Santos Costa (vitor@cos.ufrj.br)