摘要
讨论了基于惩罚域的遗传算法对神经网络结构和权值同步优化方法 .并通过实验仿真将该方法与 BP算法进行比较 ,从而验证了该方法的可行性与有效性 .
It is represented in this paper how to use hereditary algorithm based on penalty field to approach synchronousoptimizing the construction and weights of nerve net. And its feasibility as well as validity was also proved by comparing it to BP algorithm experiments.
出处
《纺织高校基础科学学报》
CAS
2002年第2期139-141,共3页
Basic Sciences Journal of Textile Universities
关键词
遗传算法
神经网络
惩罚域
同步优化
penalty field
hereditary algorithm
synchronousoptimizing
nerve net
