Proceedings:
Book One
Volume
Issue:
Proceedings of the AAAI Conference on Artificial Intelligence, 18
Track:
Student Abstracts
Downloads:
Abstract:
In this paper, I describe BN-Tools, an open-source, Java-based library for experimental research in Bayesian belief networks that implements several popular algorithms for estimation (approximate inference) and learning along with a graphical user interface for interactive display and editing of graphical models. Included in the discussion are our implementations of the Lauritzen-Spiegelhalter algorithm, an adaptive importance sampling algorithm, the K2 learning algorithm, and two genetic algorithm wrappers.
AAAI
Proceedings of the AAAI Conference on Artificial Intelligence, 18