Performance Evaluation of Multiagent Genetic Algorithm

Zhong et al. (2004 [IEEE Trans. on Systems, Man and Cybernetics (Part B), 34: 1128–1141]) proposed the multiagent genetic algorithm (MAGA) in their publication titled “A multiagent genetic algorithm for global numerical optimization”. The MAGA exploits the known characteristics of some benchmark functions to achieve outstanding results. For example, the MAGA exploits the fact that all variables have the same numerical value at the global optimum and the same upper and lower bounds to solve several 100 dimensional and 1000 dimensional benchmark problems with high precision requiring on average 7000 and 16,000 function evaluations respectively. In this paper, we evaluate the performance of the MAGA experimentally1 and demonstrate that the performance of the MAGA significantly deteriorates when the relative positions of the variables at the global optimal point are shifted with respect to the search ranges.

[1]  Lishan Kang,et al.  An Adaptive Evolutionary Algorithm for Numerical Optimization , 1996, SEAL.

[2]  Licheng Jiao,et al.  A novel genetic algorithm based on immunity , 2000, IEEE Trans. Syst. Man Cybern. Part A.

[3]  Ioannis B. Theocharis,et al.  Microgenetic algorithms as generalized hill-climbing operators for GA optimization , 2001, IEEE Trans. Evol. Comput..

[4]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[5]  Yuping Wang,et al.  An orthogonal genetic algorithm with quantization for global numerical optimization , 2001, IEEE Trans. Evol. Comput..

[6]  Petros Koumoutsakos,et al.  Learning Probability Distributions in Continuous Evolutionary Algorithms - a Comparative Review , 2004, Nat. Comput..

[7]  Krishna R. Pattipati,et al.  Sequential testing algorithms for multiple fault diagnosis , 2000, IEEE Trans. Syst. Man Cybern. Part A.

[8]  Jing J. Liang,et al.  Novel composition test functions for numerical global optimization , 2005, Proceedings 2005 IEEE Swarm Intelligence Symposium, 2005. SIS 2005..

[9]  L. Darrell Whitley,et al.  Cellular Genetic Algorithms , 1993, ICGA.

[10]  Weicai Zhong,et al.  A multiagent genetic algorithm for global numerical optimization , 2004, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[11]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .