摘要
对Pincer-Search进行了改进,提出了一种新的发现最大频繁项目集的算法NDMFS.由于NDMFS算法采用了位阵存储技术和自底向上、自顶向下双向搜索策略,从而进一步提高了算法的效率,显著降低了系统的I/O成本和CPU时间.
In this paper,Pincer- Search algorithm is improved,and proposes an new algorithm for mining maxmum frequent itemsets based on set - enumeration tree. The proposed algorithm makes enough use of bitmatrix- stored technology, and combines the bottom - up and top - down. Efficiency of algorithm is more high,cost of I/O and CPU time is reduced remarkably.
出处
《德州学院学报》
2003年第4期60-64,共5页
Journal of Dezhou University
关键词
数据挖掘
关联规则
频繁项目集
最大频繁项目集
位阵
集合枚举树
data mining
association rules
frequent itemset
maxmum frequent itemset
bitmatrix
set-enumeration tree
