基于混合遗传算法的分类规则挖掘方法研究  

Classification Rule Mining Research Based on Hybrid Genetic Algorithm

作　　者：王祥瑞[1] 

机构地区：[1]吉林建筑工程学院计算机科学与工程学院,长春130118

出　　处：《吉林建筑工程学院学报》2011年第6期68-71,共4页Journal of Jilin Architectural and Civil Engineering

基　　金：吉林省教育厅"十二五"科学技术研究项目(2011第400号)

摘　　要：基于传统遗传算法的分类规则挖掘方法,通常存在挖掘出的规则质量不高,优化后种群中的冗余规则太多,分类准确率较低等问题.文中分析了分类规则挖掘原理,提出基于混合遗传算法的分类规则挖掘方法能够有效地克服上述缺点,从而提高分类规则挖掘的准确性.Traditional classification rule mining based on genetic algorithm usually exist the problem of low quality of the rules mined,redundant rules in the population are too much after optimization and classification accuracy is inaccuracy.This paper analyses the principles of classification rules mining,and proposed classification rules mining methods based on hybrid genetic algorithm,which can effectively overcome above disadvantages and improve the accuracy of classification rule mining.

关 键 词：数据挖掘 分类规则 遗传算法 局部搜索 混合遗传算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]