一种基于矩阵优化的高效频繁项集挖掘算法  

An Efficient Mining Algorithm for Mining Frequent Itemsets Based on Matrix Optimization

作　　者：陈万敏 Chen Wan-min(Lanzhou Central Sub-branch,the People’s Bank of China,Lanzhou 730000,Gansu Province,China)

机构地区：[1]中国人民银行兰州中心支行,甘肃兰州730000

出　　处：《科学与信息化》2021年第23期43-45,共3页Technology and Information

摘　　要：本文提出一种基于布尔矩阵优化的Apriori改进算法来挖掘关联规则,通过减少对数据库的扫描次数以及减少候选项集的数量来提高算法的执行效率。通过固定事务数量和变化,来分别比较MWS_Apriori和原始算法的运行时间。实验结果表明,MWS_Apriori算法可有效减少冗余候选项集的产生,对频繁项集可快速且有效地进行挖掘。This article proposes an improved Apriori algorithm based on Boolean matrix optimization to mine association rules,which improves the implementation efficiency of the algorithm by reducing the number of database scans and the number of candidate itemsets.The running time of MWS_Apriori and original algorithm is compared by fixing the number of things and changes.Experimental results show that the MWS_Apriori algorithm can effectively reduce the generation of redundant candidate itemsets,and can quickly and effectively mine frequent itemsets.

关 键 词：关联规则 布尔矩阵 MWS_Aprioir算法 频繁项集 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]