A full variate Gaussian model-based RM-MEDA without clustering process
- 70 Downloads
A regularity model-based multi-objective estimation of distribution algorithm (RM-MEDA) is an excellent multi-objective estimation of distribution algorithm proposed in recent years. However, the performance of RM-MEDA is seriously affected by its clustering process. In order to avoid the influence of the clustering process, this paper presents a novel full variate Gaussian model-based (FGM-based) RM-MEDA without clustering process, named FRM-MEDA. In FRM-MEDA, the clustering process is removed from the original algorithm and the full variate Gaussian model (FGM) is introduced to keep the population diversity and make up the loss of the performance caused by removing the clustering process. Meanwhile, the introduction of FGM makes the FRM-MEDA faster and more stable when solving all the test instances. In addition, variable variance of FGM is presented to enhance the exploring ability of FRM-MEDA. The experiments demonstrate that the proposed algorithm significantly outperforms the RM-MEDA without clustering process and the RM-MEDA with K equal to AVE K .
KeywordsEstimation of distribution algorithm Multi-objective optimization Number of clusters Full variate Gaussian model
This work has been funded by the Project of Science and Technology for Graduate Students with No. CDJXS12180003, and the Scientific and Technological Research Program of Chongqing Municipal Education Commission under Grant No. KJ1400409.
- 2.Zhang J et al (2009) Multilayer ensemble pruning via novel multi-sub-swarm particle swarm optimization. J Univers Comput Sci 15(4):840–858Google Scholar
- 10.Laumanns M, Ocenasek J (2002) Bayesian optimization algorithms for multi-objective optimization [M]. Parallel Problem Solving from Nature-PPSN VII. Springer, Berlin, pp 298–307Google Scholar
- 11.Pelikan M, Sastry K, Goldberg DE (2005) Multi-objective hBOA, clustering, and scalability[C]. In: Proceedings of the 2005 conference on Genetic and evolutionary computation ACM, pp 663–670Google Scholar
- 12.Sastry K, Goldberg DE, Pelikan M (2005) Limits of scalability of multi-objective estimation of distribution algorithms[C]. In: Evolutionary Computation, 2005. The 2005 IEEE Congress on IEEE 3: pp 2217–2224Google Scholar
- 16.Maezawa K, Handa H (2015) Memetic algorithms of graph-based estimation of distribution algorithms [C]. In: The 18th Asia Pacific Symposium on Intelligent and Evolutionary Systems (IES 2014). Springer International Publishing 2, pp 647–656Google Scholar
- 17.Miettinen K (1999) Nonlinear multi-objective optimization [M]. Springer, BerlinGoogle Scholar
- 32.Van Veldhuizen DA, Lamont GB (2000) On measuring multi-objective evolutionary algorithm performance. In: Proceedings of the 2000 IEEE Congress on Evolutionary Computation, CEC 2000, vol 1. IEEE Service Center, Piscataway, New Jersey, pp 204–211Google Scholar