Proceedings of the 11 th Annual ERCIM Workshop on Constraint Solving and Constraint Programming

Combining machine learning [13, 16] and constraint satisfaction techniques is attracting considerable attention from a variety of researchers in the constraints community. In this talk we will review several examples of how techniques from the field of machine learning can be successfully exploited in the domain of constraint satisfaction. In particular, we will consider how we can exploit machine learning to help automatically acquire constraints [1, 2, 5, 11, 12], acquire preferences from interactions with users [18], inform the modelling and reformulation process [3, 5] and assist in constraint solving [4, 6, 8, 10, 14, 15, 17]. Finally, a number of challenges in the field will be discussed.