講座內(nèi)容： The class of CRC constraints generalizes several tractable classes of constraintsand is expressive enough to model problems in domains such as temporalreasoning, geometric reasoning, and scene labelling. This paper presents thefirst distributed deterministic algorithm for connected row-convex (CRC)constraints. Our distributed (partial) path consistency algorithm efficientlytransforms a CRC constraint network into an equivalent constraint network,where all constraints are minimal (i.e., they are the tightest constraints) andgenerating all solutions can be done in a backtrack- free manner. When comparedwith the state-of-the-art distributed algorithm for CRC constraints, which is arandomized one, our algorithm guarantees to generate a solution for satisfiableCRC constraint networks and it is applicable to solve large networks in realdistributed systems. The experimental evaluations show that our al- gorithmoutperforms the state-of-the-art algorithm in both practice and theory.This paper is a joint work with Shufeng Kong and Jae Hee Lee. It wasaccepted to the Sixteenth International Joint Conference on Autonomous Agentsand Multi-agent Systems to be held in Sao Paulo, Brazil in 8-12 May, 2017.