A Software for Learning Bayesian Networks

How to use URLearing?

This is software we have developed for learning a Bayesian network structure for a dataset which optimizes the score for a dataset. The C++ source code is available as a netbeans project. It requires Boost (C++ source libraries). This file describes the necessary steps to configure netbeans to compile the code. This file describes the necessary steps to add a new configuration (algorithm) to the project.

You can also clone the developer version of the code with this command:
git clone git@bitbucket.org:bmmalone/urlearning-cpp.git

Furthermore, a set of scripts are available which implement most of the developments we published in UAI 2014 and AAAI 2014. They are available from the following repository:
git clone git@bitbucket.org:bmmalone/urlearning-python.git
The readme.uai2014.txt file describes the prerequisites and necessary steps to run the code.

There is also an older Java version available. It is not as efficient as the C++ version, but it includes more features and does not require any external dependencies.