DocumentCode :
2805261
Title :
Modified evolution strategies with a diversity-based parent-inclusion scheme
Author :
Huang, Tao-Yuan ; Chen, Yung-Yaw
Author_Institution :
Inst. of Electr. Eng., Nat. Taiwan Univ., Taipei, Taiwan
fYear :
2000
fDate :
2000
Firstpage :
379
Lastpage :
384
Abstract :
There are two major variants of evolution strategies: the (μ, λ) evolution strategy and the (μ+λ) evolution strategy. Back et al., in a survey paper (1997), suggested that preference of either strategy is most likely problem-dependent, and that both strategies are special cases of the general (μ, κ, λ) evolution strategy, where κ denotes the maximum life span of an individual. In this paper, a diversity-based fuzzy parent-inclusion scheme is devised to make compromises between both strategies and thus retain advantages from each. Simulation results provided do manifest the virtue of this modified evolution strategy
Keywords :
evolutionary computation; fuzzy set theory; diversity-based fuzzy parent-inclusion scheme; maximum life span; modified evolution strategies; problem-dependent strategy preference; Evolution (biology); Evolutionary computation; Flowcharts; Fuzzy control; Genetic algorithms; Genetic mutations; Genetic programming; Optimization methods; Random number generation; Search methods;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Control Applications, 2000. Proceedings of the 2000 IEEE International Conference on
Conference_Location :
Anchorage, AK
Print_ISBN :
0-7803-6562-3
Type :
conf
DOI :
10.1109/CCA.2000.897454
Filename :
897454
Link To Document :
بازگشت