Portfolio approaches in constraint programming
                               Invited Talk

                              Roberto Amadini

               University of Melbourne, Melbourne, Australia



   Abstract. Recent research has shown that the performance of a single,
   arbitrarily efficient algorithm can be significantly outperformed by us-
   ing a portfolio of —possibly on-average slower— algorithms. Within the
   Constraint Programming (CP) context, a portfolio solver can be seen
   as a particular constraint solver that exploits the synergy between the
   constituent solvers of its portfolio for predicting which is (or which are)
   the best solver(s) to run for solving a new, unseen instance.
   In the work we examined the benefits of portfolio solvers in CP. We
   focused in particular on sequential approaches, i.e., portfolio solvers al-
   ways running on a single core. We started from a first empirical evalua-
   tion on portfolio approaches for solving Constraint Satisfaction Problems
   (CSPs), and then we improved on it by introducing new data, solvers,
   features, algorithms, and tools. Afterwards, we addressed the more gen-
   eral Constraint Optimization Problems (COPs) by implementing and
   testing a number of models for dealing with COP portfolio solvers. Fi-
   nally, we have come full circle by developing sunny-cp: a sequential CP
   portfolio solver that turned out to be competitive also in the MiniZinc
   Challenge, the reference competition for CP solvers.