This package supports learning discrete Bayesian networks from continuous data, following our work in Learning Discrete Bayesian Networks from Continuous Data, which has been submitted to the Journal of Artificial Intelligence Research (JAIR).
The api is:
- Discretize_All(data,graph,continuous_index,u_cycle) : discretize continuous variables in a Bayesian network for which the network structure is known in advance
- Learn_DVBN(data,continuous_index,order,u_parent,u_cycle): learn a discrete Bayesian network given a topological ordering of the variables
- Learn_Discrete_Bayesian_Net(data,continuous_index,u_parent,u_cycle,times): learn a discrete Bayesian network for which the network structure is not known in advance.