摘要
关联规则是数据挖掘中一个重要课题.文章给出一种基于遗传算法和蚂蚁算法相结合的多维关联规则挖掘算法.新算法利用了遗传和蚂蚁算法共有的良好全局搜索能力,并克服了遗传算法局部搜索能力弱和蚂蚁算法搜索速度慢的缺陷.实验结果表明,新算法在对具有稀疏特性的多维关联规则的挖掘中体现了良好的性能.
Association rules mining is very important in the application of data mining. In this paper,a method of mining Multidi mensional Association Rule is proposed-based on the combination of genetic algorithm and ant algorithm. The new algorithm has the outstanding capacity for global searching. It overcomes the weakness in local searching of the inheritance algorithm and the slowness of ant algorithm. The experimental results show that this new algorithm has proven its significant performance in the sparse multidimensional association rule mining.
出处
《小型微型计算机系统》
CSCD
北大核心
2006年第2期291-294,共4页
Journal of Chinese Computer Systems
基金
国家"八六三"计划基金项目(2003AA412020)资助.
关键词
遗传算法
蚂蚁算法
关联规则
数据挖掘
genetic algorithm ant algorithm
association rule data mining