摘要
在简单遗传算法基础上,引入“相似度”和“杂交优势”思想,将原来的交叉和变异两个算子合二为一,提出了一个新的遗传算子:基于相似度的交叉变异算子。将算子应用于分类规则的挖掘中,既可提高算法的收敛速度,又能抑制早熟收敛现象的发生。
Based on the simple genetic algorithm the idea of ＂similarity degree＂ and ＂beneficial crossover＂ is introduced and a new operator called similarity-based crossover-mutation is presented by combining the crossover with the mutation. Applying new operator to mining classification rules it is found to speed up the algorithm convergence and restrain the premature convergence to some extent.
出处
《成都信息工程学院学报》
2007年第1期64-68,共5页
Journal of Chengdu University of Information Technology
关键词
数据挖掘
遗传算法
相似度
交叉变异算子
分类规则
data mining
genetic algorithm
similarity degree
crossover-mutation operato4r
classification rule
