A Markov network is an undirected graph of nodes representing variables and edges representing dependencies amongst these variables. Each clique in this undirected graph represents a set of dependent or possibly dependent variables, and has associated with it a potential function from the set of all assignments to the variables to the nonnegative real numbers. It is similar to a Bayesian network in its representation of dependencies, but a Markov network can represent dependencies that a Bayesian network can not.

The potential functions used in a Markov network do not necessarily have a probabilistic interpretation by themselves, but a higher value indicates a more probable assignment to the variables in a clique. The network is used to represent the joint distribution over all variables represented by nodes in the graph. Two variables are conditionally independent if there is no edge between the nodes representing them in the graph.