Reducing search space in optimization problems with data mining techniques

Finding the optimal global solution in optimization problems is such an important issue that various related approaches have been proposed so far. An effective attempt before solving such problems is to reduce the search space in such a way that the search is concentrated in a smaller subspace and therefore the probability of finding the optimal global solution increases. In this article, three methods of clustering, classification and association in data mining are used to reduce the search space in a nonlinear optimization problem. After that, using the Genetic Algorithm, the problem is solved on the entire initial feasible space and the reduced spaces resulting from three data mining methods. The results show that by combining data mining methods and Genetic Algorithm, more accurate approximations for the global optimal solution of the problem can be obtained.