Turning High-Dimensional Optimization Into Computationally Expensive Optimization

Divide-and-conquer (DC) is conceptually well suited to deal with high-dimensional optimization problems by decomposing the original problem into multiple low-dimensional subproblems, and tackling them separately. Nevertheless, the dimensionality mismatch between the original problem and subproblems makes it nontrivial to precisely assess the quality of a candidate solution to a subproblem, which has been a major hurdle for applying the idea of DC to nonseparable high-dimensional optimization problems. In this paper, we suggest that searching a good solution to a subproblem can be viewed as a computationally expensive problem and can be addressed with the aid of meta-models. As a result, a novel approach, namely self-evaluation evolution (SEE) is proposed. Empirical studies have shown the advantages of SEE over four representative compared algorithms increase with the problem size on the CEC2010 large scale global optimization benchmark. The weakness of SEE is also analyzed in the empirical studies.