Max-Product Belief Propagation for Linear Programming: Applications to Combinatorial Optimization
This work provides a theoretical generalization for using belief propagation in combinatorial optimization, addressing a gap in existing methods that were limited to specific problem setups.
The authors derived a generic convergence criterion for max-product belief propagation when solving linear programming formulations, and demonstrated its applicability to classical combinatorial optimization problems such as maximum weight perfect matching and traveling salesman.
The max-product {belief propagation} (BP) is a popular message-passing heuristic for approximating a maximum-a-posteriori (MAP) assignment in a joint distribution represented by a graphical model (GM). In the past years, it has been shown that BP can solve a few classes of linear programming (LP) formulations to combinatorial optimization problems including maximum weight matching, shortest path and network flow, i.e., BP can be used as a message-passing solver for certain combinatorial optimizations. However, those LPs and corresponding BP analysis are very sensitive to underlying problem setups, and it has been not clear what extent these results can be generalized to. In this paper, we obtain a generic criteria that BP converges to the optimal solution of given LP, and show that it is satisfied in LP formulations associated to many classical combinatorial optimization problems including maximum weight perfect matching, shortest path, traveling salesman, cycle packing, vertex/edge cover and network flow.