URLearning ("You are learning"): URL Software Package for Learning Optimal Bayesian Networks

Brandon Malone, Changhe Yuan

Abstract

URLearning is a Java software package that implements the search algorithms developed by the URL Lab (both current members and alumni) for learning optimal Bayesian networks. A Windows installable software package has been created so that you can use our algorithms directly in the popular machine learning software -- Weka. Finally, we have shifted to develop C++ versions of the algorithms for more consistent run performance across platforms; the C++ source code is available for download here. Users of our software can cite one or more of the following papers that describe the algorithms when you see fit. Disclaimer: The software is provided as is without any guarantee, and can be reused and redistributed except for commercial use.

Download Java software (size: 4.4M; updated: 4/08/2013)

Download Weka package (size: 18M; updated: 6/18/2013)

C++ source code (size: 2.83M; updated: 8/24/2014)

Bibtex

@INPROCEEDINGS{Fan14learning,
author = {Xiannian Fan, Brandon Malone, Changhe Yuan},
title = {Learning Optimal Bayesian Network Structures with Constraints Learned from Data},
booktitle = {In Proceedings of the 30th Conference on Uncertainty in Artificial Intelligence (UAI-14)},
address = {Quebec City, Quebec, Canada},
pages = {200--209},
year = {2014}
}

@INPROCEEDINGS{Fan14Tightening,
author = {Xiannian Fan, Changhe Yuan, Brandon Malone},
title = {Tightening Bounds for Bayesian Network Structure Learning},
booktitle = {In Proceedings of the 28th AAAI Conference on Artificial Intelligence (AAAI-14)},
address = {Quebec City, Quebec, Canada},
pages = {2439--2445},
year = {2014}
}

@article{Yuan13learning,
author = {Changhe Yuan and Brandon Malone},
title = {Learning Optimal Bayesian Networks: A Shortest Path Perspective},
booktitle = {Journal of Artificial Intelligence Research (JAIR)},
year = {2013},
pages = {23--65},
volume = {48}
}

@INPROCEEDINGS{Malone13evaluating,
author = {Brandon Malone and Changhe Yuan},
title = {Evaluating Anytime Algorithms for Learning Optimal Bayesian Networks},
booktitle = {In Proceedings of the 29th Conference on Uncertainty in Artificial Intelligence (UAI-13)},
address = {Seattle, Washington},
year = {2013}
}

@INPROCEEDINGS{Yuan12improved,
author = {Changhe Yuan and Brandon Malone},
title = {An Improved Admissible Heuristic for Learning Optimal Bayesian Networks},
booktitle = {Proceedings of the 28th Conference on Uncertainty in Artificial Intelligence (UAI-12)},
year = {2012},
address = {Catalina Island, CA}
}

@INPROCEEDINGS{Yuan11learning,
author = {Changhe Yuan and Brandon Malone and Xiaojian Wu},
title = {Learning Optimal {B}ayesian Networks Using {A}* Search},
booktitle = {Proceedings of the 22nd International Joint Conference on Artificial Intelligence (IJCAI-11)},
year = {2011},
pages = {2186--2191},
address = {Helsinki, Finland}
}

@INPROCEEDINGS{Malone11memory,
author = {Brandon Malone and Changhe Yuan and Eric A. Hansen},
title = {Memory-Efficient Dynamic Programming for Learning Optimal {B}ayesian Networks},
booktitle = {Proceedings of the 25th AAAI Conference on Artificial Intelligence (AAAI-11)},
year = {2011},
pages = {1057--1062},
address = {San Francisco, CA}
}

@INPROCEEDINGS{Malone11improving,
author = {Brandon Malone and Changhe Yuan and Eric Hansen and Susan Bridges},
title = {Improving the Scalability of Optimal {B}ayesian Network Learning with Frontier Breadth-First Branch and Bound Search},
booktitle = {Proceedings of the 27th Conference on Uncertainty in Artificial Intelligence (UAI-11)},
year = {2011},
pages = {479--488},
address = {Barcelona, Catalonia, Spain}
}