摘要
通过改变遗传规划算法中初始群体的生成方法,改变变异策略和修正适应度函数,对遗传规划算法进行了改进,并通过符号回归数值实验对改进后算法的性能进行了测试,且将改进后的算法与改进前以及其它改进算法进行了比较,数值实验结果表明,改进后的算法有效地提高了遗传规划的效率。
To improve genetic programming algorithm, a new method for creating initial population is proposed and mutation-probability and fitness function are adjusted. Numerical experiments of symbolic regression given in the third part show that the improved algorithm in this paper is feasible and more efficient than the traditional one and some other improved ones.
出处
《运筹与管理》
CSCD
2007年第1期67-70,共4页
Operations Research and Management Science
关键词
运筹学
遗传规划
改进算法
符号回归
operations research
genetic programming
improved algorithm
symbolic regression
