During the last decades, significant improvements have been achieved for solving complex combinatorial optimization problems issued from real world applications. To tackle large scale instances and intricate problem structures, sophisticated solving techniques have been developed, combined, and hybridized to provide efficient solvers. Combinatorial problems are often modeled as Constraint Satisfaction Problems or constraint optimization problems, which consist of a set of variables, a set of possible values for these variables and a set of constraints to be satisfied. However, solvers or hybridization of solvers become more and more complex: the user must select various solving and hybridization strategies and tune numerous parameters. Moreover, it is well-known that an a priori decision concerning strategies and parameters is very difficult since strategies and parameters effects are rather unpredictable and may change during solving.