VMODE: A HYBRID METAHEURISTIC FOR THE SOLUTION OF LARGE SCALE OPTIMIZATION PROBLEMS

Abstract

Large scale continuous optimization problems have become increasingly common in real-world problems. The resolutions of these are computationally expensive, so the use of scalable and efficient algorithms is of particular interest. In this paper is proposed a hybrid algorithm, VMODE, which results from the combination of DE algorithm, known for its simplicity and efficiency and VMO, a population-based algorithm with encouraging results in continuous optimization. A comparison among the three algorithms is done using the 15 proposed functions for CEC-2013 (Special Session and Competition on Large-Scale Global Optimization) demonstrating the superiority of the algorithm VMODE