Machine learning for constraint acquisition

To use constraint programming, one needs to formulate a model that consists of a set of constraints. Acquiring and formulating these constraints is a non-trivial task, in which one typically has to interact and collaborate intensively with a domain expert. An alternative for eliciting the constraints from the expert is to use possibly available data sets in order to automatically learn the constraints or mine for patterns that could help identify the right constraints. This tutorial will survey a number of basic machine learning methods for learning constraints, it will make the link to computational learning theory and will also use this to position existing work on automated learning of constraints in this context.Tutorialstatus: publishe