Computational experiments are performed with the simple genetic algorithm (SGA) for a wide range of intron lengths, crossover rates, mutation rates, and replication rates. The best-adapted species emerges widely with a critical intron length proportional to the overall crossover rate in an environment of discontinuous evolution (quasi-macroevolution). This result coincides with the relation between the intron length and mating rate observed in actual Eukaryotes. On the basis of this knowledge, a method for optimizing the crossover rate in the SGA is proposed for the purpose of accomplishing artificial macroevolution in engineering problems.
|ジャーナル||JSME International Journal, Series C: Dynamics, Control, Robotics, Design and Manufacturing|
|出版ステータス||Published - 1998 9|
ASJC Scopus subject areas