DUCT: An Upper Confidence Bound Approach to Distributed Constraint Optimization Problems

The Upper Confidence Bounds (UCB) algorithm is a well-known near-optimal strategy for the stochastic multi-armed bandit problem. Its extensions to trees, such as Tree (UCT) algorithm, have resulted in good solutions problem of Go. This paper introduces DUCT, distributed inspired by UCT, solving Distributed Constraint Optimization Problems (DCOP). on solution quality are provided, and experiments show that, compared existing DCOP approaches, DUCT able solve very large problems much more efficiently, or find significantly higher solutions.