PowerBayes 1.0 is a package for structure learning of Bayesian networks. It contains implementations of many common structure-learning algorithms and new algorithms using constraint satisfaction for learning models with improved structural accuracy.

See “Learning the Structure of Bayesian Networks with Constraint Satisfaction.” Fast, Andrew. Ph.D. Thesis, University of Massachusetts Amherst. (2009) for additional information on the algorithms appearing in PowerBayes.

PowerBayes 1.0 is written in Java. The PowerBayes distribution includes:

  • Java source code
  • All necessary libraries
  • README file containing instructions for running structure learning algorithms

Download PowerBayes 1.0 distribution

Download Bayesian network models for use with PowerBayes

PowerBayes is designed and implemented by the Knowledge Discovery Laboratory in the College of Information and Computer Sciences at the University of Massachusetts Amherst.