LDecode：具有线性复杂度的GEP适应度评价算法

基因表达式编程(Gene Expression Programming,GEP)在处理复杂长基因时的空间、时间效率较低,为解决这一问题,提出并实现了具有线性复杂度的染色体适应度评价算法.分析了传统GEP算法中借助ET(Expression Tree)树进行染色体适应度评价的局限性;提出并实现了具有线性复杂度的染色体适应度评价算法LDecode算法;针对染色体长度、种群大小、测试数据集大小、进化代数等不同参数,对提出的染色体适应度评价算法进行了评价和分析.试验表明,提出的评价算法运行速度较传统基于ET树的GEP提高了4.5～5.1倍,时间、空间复杂度均为O(n).

LDecode:A Novel Decoding Algorithm on Gene Expression Programming with Linear Complexity

The efficiency of gene expression programming (GEP) is low in processing complex gene with large length. In order to solve the problem, a novel algorithm with linear complexity was proposed. The main contributions include: 1) Analyzing the limit in conventional gene expression programming, 2) Proposing a novel LDecode algorithm to evaluate the fitness of chromosome speedy.The extensive experiments demonstrated that this algorithm is faster than traditional one by 4.5~5.1 times on different parameters, and the time complexity and space complexity are linear.