基于GEP基因约简策略的多项式因式分解  

FactorGEP:a novel factorization of polynomial algorithm based on GEP with gene reduction strategy

作　　者：陈瑜[1] 唐常杰[1] 朱明放[1,2] 代术成[1] 朱锐[3] 姜玥[1,4] 李川[1] 

机构地区：[1]四川大学计算机学院,成都610065 [2]江苏技术师范学院计算机科学与工程学院,江苏常州213001 [3]西安财经学院信息与教育技术中心,西安710061 [4]西南民族大学计算机科学与技术学院,成都610041

出　　处：《吉林大学学报（工学版）》2009年第6期1612-1617,共6页Journal of Jilin University:Engineering and Technology Edition

基　　金：国家自然科学基金项目(60473071,60773169);“十一五”国家科技支撑计划项目(2006BAI05A01)

摘　　要：将基于基因表达式编程(GEP)的公式发现技术用于多项式因式分解。对标准GEP中个体染色体适应度评价算法和遗传算子进行了改进,提出了一种新的基于GEP的基因约简算法及多项式因式发现算法(以下称FactorGEP算法),分析了多项式因式发现的特殊困难,提出了将有效基因个数作为适应度评价因素的解决方案。试验结果表明,FactorGEP算法完全不需要关于分解的多项式的任何先验知识,可以自动对多项式进行因式分解,并找到最佳的分解。A novel factorization of polynomial algorithm was proposed based on Gene Expression Programming（GEP）.The fitness evaluation method for individual chromosome and the genetic operator of traditional GEP were improved.The difficulty in factorization of polynomial was analyzed.Then a novel factorization of polynomial algorithm based on gene reduction in GEP,FactorGEP,was proposed.In the proposed algorithm the number of valid genes in chromosome is taken as a factor in the fitness evaluation.Experimental results show that the FactorGEP algorithm can automatically perform factorization of polynomial well without any priori knowledge of the polynomial,and it can find the optimal factorization.

关 键 词：人工智能 基因表达式编程 多项式因式分解 进化计算 适应度计算 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]