Home

Here we will attempt to incorporate Bayesian Network learning and inference algorithms into the Neo4J graph database. The interest is in scalable, database-backed Bayesian reasoning. The substrate is Neo4j because property graphs are a more convenient and potentially more efficient metaphor for Bayesian Networks than relational databases.

Our first task is to tie existing tools to Neo4J. From there, we will explore some of the cool advances in scalable Bayesian reasoning that database-backed algorithms can offer.

Collaborators are very welcome. Feel free to contribute to the work or the references page.