Optimization of Noisy Fitness Functions by means of Genetic Algorithms using History of Search
Yasuhito Sano, Hajime Kita · IEEJ Transactions on Electronics Information and Systems · 2002
This paper discusses optimization of functions with uncertainty by means of Genetic Algorithms (GA). For such problems, there have been proposed methods of sampling fitness values several times and taking average of them for evaluation of each individual. However, in important applications having uncertain fitness functions such as online adaptation of real systems and optimization through complicated computer simulation using random variables, possible number of fitness evaluation is quite limited. Hence, methods achieving optimization with less number of fitness evaluation is needed. In the present paper, the authors propose a GA utilizing history of search (Memory-based Fitness Evaluation GA: MFEGA) so as to reduce the number of fitness evaluation for such applications of GA. In the MFEGA, value of fitness function at a novel search point is estimated not only by the sampled fitness value at that point but also by utilizing the fitness values of individuals stored in the history of search. Numerical experiments show that the proposed method outperforms the conventional GA of sampling fitness values several times at each search point in noisy environment.