The moral graph of a Directed acyclic graph joins every pair of vertices with a common child and then replaces every directed edge by an undirected edge.
For a strictly positive probability density, factorization into clique potentials of an undirected graph is equivalent to the global Markov conditional-independence property for that graph.
Articles by others on the same topic
A **moral graph** is a concept used in the fields of graph theory and probabilistic graphical models, particularly in the context of Bayesian networks and Markov networks. The moral graph is derived from a directed acyclic graph (DAG) representing a Bayesian network. ### How to Construct a Moral Graph: 1. **Start with a Directed Graph:** Begin with a Bayesian network, which is typically represented as a directed acyclic graph (DAG).