Title :
A novel adaptive genetic algorithms
Author :
Liu, De-peng ; Feng, Shu-ting
Author_Institution :
Sch. of Sci., Hangzhou Inst. of Electron. Eng., China
Abstract :
This work presents a modified genetic algorithm that is based on the tuning of the mutation probability by the value of individual fitness. The fine modular in current generation is easy to survive in the offspring, and at the same time, the variety of the population is also guaranteed. In the modified scheme, the order of crossover and mutation is changed in order to avoid the repetition in the computation of individual fitness. Simulation result have shown that the modified scheme is prior to the GAs commonly used.
Keywords :
genetic algorithms; probability; adaptive genetic algorithm; individual fitness; mutation probability; Algebra; Algorithm design and analysis; Cybernetics; Genetic algorithms; Genetic mutations; Least squares methods; Optimization methods; Registers; Turning;
Conference_Titel :
Machine Learning and Cybernetics, 2004. Proceedings of 2004 International Conference on
Print_ISBN :
0-7803-8403-2
DOI :
10.1109/ICMLC.2004.1380721
