摘要
针对遗传算法的有效性一般是通过数值实验来说明这一问题,本文在三个假设的条件下给出了一类混合遗传算法依概率收敛的证明,而且得到了在此算法的框架下通过改变子种群Ⅱ的变异算子而不改变依概率收敛特性的结论.
For the problem that the effectivenes of Genetic Algorithm is shown generally by numerical experiments, the convergence in probability of a class of hybrid genetic Algorithm based on the condition of three hypothesizes is proved in the paper and the conclusion is drawn that the convergence in probability is not unchanged if mutation operator of sub-population Ⅱ is changed under the algorithm frame.
出处
《数学的实践与认识》
CSCD
北大核心
2006年第12期169-173,共5页
Mathematics in Practice and Theory
基金
陕西省教育厅专项资金项目(05JK303)
关键词
混合遗传算法
收敛性证明
依概率
hybrid Genetic Algorithm
convergence proof
in probability