Recursive Best-First AND/OR Search for Optimization in Graphical Models

The paper presents and evaluates the power of limited memory best-first search over AND/OR spaces for optimization tasks in graphical models. We propose Recursive Best-First AND/OR Search with Overestimation (RBFAOO), a new algorithm that explores the search space in a best-first manner while operating with restricted memory. We enhance RBFAOO with a simple overestimation technique aimed at minimizing the overhead associated with re-expanding internal nodes and prove correctness and completeness of RBFAOO. Our experiments show that RBFAOO is often superior to the current state-of-the-art approaches based on AND/OR search, especially on very hard problem instances.