摘要
关联规则是数据挖掘中一个重要的研究方向。文章提出了一种基于遗传算法的多维多层关联规则挖掘算法。算法针对现有多维多层关联规则挖掘中存在阈值定义不合理的缺点,依据多维和多层数据的共同特点,提出了一种启发式的阈值自定义方式,来提高挖掘效率和结果的准确性。实验结果证明这种算法是有效。
Association rules mining is very important in the application of data mining.In this paper,a method of mining multi-dimensional and multi-level Association Rule was proposed based on the genetic algorithm.It put forward a heuristic user-defined method,which based on the common characteristic of multi-dimensional And multi-level data, to overcome the drawbacks caused by the unreasonable method of defining threshold.So the precision and efficiency of mining association rules is improved.The experimental results show that the efficiency of the algorithm for large databases.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第12期174-176,216,共4页
Computer Engineering and Applications
基金
国家863高技术研究发展计划资助项目(编号:2003AA412020)
关键词
多维多层
关联规则
遗传算法
数据挖掘
multidimensional and multiple-level, association rule,Genetic Algorithm,Data Mining
