Parallel Recursive Best-First AND/OR Search for Exact MAP Inference in Graphical Models

Akihiro Kishimoto, Radu Marinescu, Adi Botea

Neural Information Processing Systems 

The paper presents and evaluates the power of parallel search for exact MAP inference in graphical models. We introduce a new parallel shared-memory recursive best-first AND/OR search algorithm, called SPRBFAOO, that explores the search space in a best-first manner while operating with restricted memory. Our experiments show that SPRBFAOO is often superior to the current state-of-the-art sequential AND/OR search approaches, leading to considerable speed-ups (up to 7-fold with 12 threads), especially on hard problem instances.