An Improved Multi-Objective Evolutionary with Adaptable Parameters

Genetic Algorithms, introduced by Holland in 1975, are general-purpose heuristic search algorithms that mimic the evolutionary process in order to find the fittest solutions. The algorithms have received growing interest due to their ability to discover good solutions quickly for complex searching and optimization problems. The traditional GAs then have been converted to multi-objective GAs to solve multi-objective optimization problems successfully. However, GAs require parameter tunings (such as population size, mutation and crossover probabilities, selection rates) in order to achieve the desirable solutions. The task of tuning GA parameters has been proven to be far from trivial due to the complex interactions among the parameters. The objective of this research is to develop the elitist Non-dominated Sorting GA (NSGA-II) for multi-objective optimization as a parameter-less multi-objective GA. The research then will evaluate and discuss the performance of the parameter-less NSGA-II against other GAs with optimal parameter settings using the experiment result on a test problem borrowed from the literature.

Autorentext

Dr. Tran earned his Ph.D. in Computer and Information Sciences from Nova Southeastern University in Florida, M.S. degree in Computer Science from California State University at Fullerton, and B.S. degree in Information and Computer Science from University of California at Irvine. Currently, he is an adjunct faculty and software consultant.