I'm currently working in the Probabilistic Machine Learning group
at Aalto University.
BEANDisco is software for Bayesian network structure discovery from given data. It takes discrete data as input and outputs the estimated probability for each potential arc. The probabilities can be either approximated or computed exactly (if the number of variables is small).
SLL is software for score based local structure discovery in Bayesian networks. This includes finding neighbors and Markov blankets for given target nodes. In addition, the software implements two different local-to-global structure learning algorithms for global learning.
Treedy is a heuristic for approximate weighted counting of subsets. This C++ implementation includes both Treedy and Sorted heuristics as well as the Exact algorithm for the weighted counting task.
Last modified 16 February 2018