基因表达式编程算法的改进

发布时间：2018-01-19 05:05

本文关键词： 基因表达式编程算法自适应进化分层模型 Spark分布式计算预测　出处：《计算机工程与设计》2017年12期 　论文类型：期刊论文

【摘要】：为解决基因表达式编程算法(gene expression programming,GEP)存在的收敛速度慢、早熟、易陷入局部极值点等问题,提出3个改进方法。设计自适应进化参数,实现根据进化代数和个体适应度值在群体中所处的排名,动态调整重组率和变异率;将种群按年龄分层繁衍,进一步扩大基因搜索空间并避免早熟;将GEP移植到Spark分布式框架,进行并行计算,使算法能在较短时间内处理大量搜索任务。实验结果表明,相比传统GEP,改进后的算法有更快的收敛速度、更高预测精度和稳定性。
[Abstract]:In order to solve the problems such as slow convergence rate, premature convergence and easy to fall into local extremum in gene expression programming algorithm. Three improved methods are proposed. The adaptive evolutionary parameters are designed to dynamically adjust the recombination rate and variation rate according to the rank of evolutionary algebra and individual fitness in the population. Multiply the population by age to further expand the genetic search space and avoid precocity; The GEP is transplanted to the Spark distributed framework, and parallel computation is carried out, so that the algorithm can deal with a large number of search tasks in a relatively short time. The experimental results show that compared with the traditional GEP. The improved algorithm has faster convergence speed, higher prediction accuracy and higher stability.
【作者单位】：北京工业大学计算机学院;
【分类号】：TP18
【正文快照】： 0引言基因表达式编程算法作为遗传算法和遗传编程的继承与发展[1],综合了二者的优点,在解决复杂问题上比传统的遗传算编程算法高出2-4个数量级的效率,在多种领域都有实际应用,展示了其潜力和无限前景[2-4],但是还存在很多问题需要进一步的研究和改进。王丽萍等提出了基于多偏

【相似文献】