摘要
挖掘关联规则是数据挖掘中一个重要的课题,产生频繁项目集是其中的一个关键步骤。文章提出了一种基于向量和矩阵的挖掘算法AVM,并将该算法与两种经典的发现频繁项目集的算法进行了比较。该算法只需要对数据库扫描一遍,并且存放辅助信息所需要的空间也少。实验表明与原先的算法相比,该算法的效率较好。
Mining association rules is an important problem in data mining.Generating large itemsets is its key.This pa-per presents a novel algorithm based on vectors and matrix for finding frequent itemsets,and compares it with two tra-ditional algorithms.AVM only needs scan the database one time ,and occupies few memory for assistant information.Ex-periment results indicate that the new algorithm has good efficiency compared with presented ones.
出处
《计算机工程与应用》
CSCD
北大核心
2004年第12期170-173,共4页
Computer Engineering and Applications
关键词
数据挖掘
关联规则
频繁项集
基于向量和矩阵的算法
Data mining,Association rules,Large itemsets,An algorithm based on vectors and matrix
