Farsighted Probabilistic Sampling: A General Strategy for Boosting MaxSAT Local Search Solvers

Local search has been demonstrated as an efficient approach for both Partial MaxSAT (PMS) and Weighted PMS (WPMS), denoted as (W)PMS, two practical generalizations to the typical combinatorial problem of MaxSAT. In this work, we observe that most (W)PMS local search solvers usually flip a single variable per iteration. Such a mechanism may lead to relatively low-quality local optimal solutions, and may limit the diversity of the search directions to escape from local optima. To this end, we propose a general strategy called farsighted probabilistic sampling (FPS) to replace the single flipping mechanism to boost the (W)PMS local search algorithms. FPS considers the benefit of continuously flipping a pair of variables, so as to find higher-quality local optimal solutions. Moreover, FPS presents an effective approach to escape from local optima by preferring the best to flip among the best sampled single variable and the best sampled variable pair. Extensive experiments demonstrate that our proposed FPS strategy significantly improves the (W)PMS state-of-the-art (local search) solvers, and FPS has an excellent generalization to various (Max)SAT local search solvers.