Verifying the Effectiveness of an Evolutionary Approach in Solving Many-Objective Optimization Problems
Most approaches in the evolutionary multi-objective optimization were found to be vulnerable in solving many-objective optimization problems (four or more objectives). This is mainly due to the fact that these algorithms lack from ability to handle more than three objectives adequately. For this reason, researchers have been focusing in developing algorithms capable of addressing many-objective optimization problems. Recently, authors of the Non-dominated Sorting Genetic Algorithm-II (NSGA-II) have proposed an extension of this approach, called NSGA-III. This new approach has opened new directions for research and development to solve many-objective optimization problems. In this algorithm the maintenance of diversity among population members is aided by supplying a number of well-spread reference points. In this work, a comparative study of the performance of NSGA-II and NSGA-III was carried out. Our aim is to verify the effectiveness of NSGA-III to deal with many-objectives problems and extend the range of problems that this approach can solve. For this, the comparison was made addressing the project portfolio problem, using instances with three and nine objectives.
KeywordsPareto Front Parent Population Pareto Dominance Offspring Population Orthogonal Distance
This work was partially financed by CONACYT, COTACYT, DGEST, TECNM and ITCM.
- 1.Deb, K.: Multi-Objective Optimization using Evolutionary Algorithms, vol. 16. Wiley, New York (2001)Google Scholar
- 3.Talbi, E.G.: Metaheuristics: from Design to Implementation, vol. 74. Wiley, New York (2009)Google Scholar
- 4.Deb, K., Jain, H.: An evolutionary many-objective optimization algorithm using reference-point based nondominated sorting approach, part i: solving problems with box constraints. In: Proceedings of IEEE Transactions on Evolutionary Computation (2013)Google Scholar
- 5.Pareto, V.: Politique, Cours D’ economie. Rouge, Lausanne, Switzerland (1896)Google Scholar
- 6.Zitzler, E., Thiele, L.: Multiobjective optimization using evolutionary algorithms: a comparative case study. In: Eiben, A.E., Bäck, T., Schoenauer, M., Schwefel, H.-P. (eds.) PPSN. Lecture Notes in Computer Science, vol. 1498, pp. 292–301. Springer, Berlin (1998)Google Scholar
- 7.Zitzler, E.: Evolutionary algorithms for multiobjective optimization: methods and applications. Ph.D. dissertation, Swiss Federal Inst. Technology (ETH) Zurich, Switzerland (1999)Google Scholar
- 8.Laumanns, M., Zitzler, E., Thiele, L.: A unified model for multi-objective evolutionary algorithms with elitism. In: Proceedings of the 2000 Congress on Evolutionary Computation, 2000, vol. 1, pp. 46–53. IEEE (2000)Google Scholar
- 11.Deb, K., Agrawal, S., Pratap, A., Meyarivan, T.: A fast elitist non-dominated sorting genetic algorithm for multi-objective optimization: NSGA-II. Lect. Notes Comput. Sci. 1917, 849–858 (2000)Google Scholar
- 13.Kukkonen, S., Deb, K.: Improved pruning of non-dominated solutions based on crowding distance for bi-objective optimization problems. In: Proceedings of the World Congress on Computational Intelligence (WCCI-2006), pp. 1179–1186. IEEE Press, Vancouver (2006)Google Scholar
- 15.Castro, M.: Development and implementation of a framework for I&D in public organizations. Master’s thesis, Universidad Autónoma de Nuevo León, Mexico (2007)Google Scholar
- 16.Garcia, R.: Hyper-Heuristic for solving social portfolio problem. Master’s Thesis, Instituto Tecnológico de Cd. Madero (2010)Google Scholar