摘要
通过对关联规则挖掘技术及经典算法Apriori和FP-growth的研究和分析,提出了一种改进的频繁项集挖掘算法。该算法利用矩阵存储数据,并结合矩阵运算求项集的支持数,有效减少了事务数据库的扫描次数;利用有序频繁项目邻接矩阵创建频繁模式树,有效减少了频繁模式树的分支和层数。通过实例分析了频繁项集的挖掘过程。
In view of the association rule mining technology and the research and analysis of its classic Apriori algo- rithm and FP-growth algorithm, an advanced frequent itemsets mining algorithm is proposed. The improved algo- rithm stores database using of matrix and calculates itemsets＇ support number in terms of the matrix operation, which reduces the number of times for database scanning. The algorithm creates frequent pattern tree using of orderly frequent item adjacency matrix, which effectively reduces the branch and layer of the tree. Finally the examples ana- lyze the frequent itemsets of mining process.
出处
《计算机工程与应用》
CSCD
2012年第19期119-121,144,共4页
Computer Engineering and Applications
基金
国家自然科学基金(No.41171341)
教育部新世纪优秀人才支持计划(No.NCET-09-0126)
河南省科技创新杰出青年基金(No.114100510006)
河南省重点科技攻关计划项目(No.112102210024)
航空科学基金(No.2010ZG55029)
关键词
数据挖掘
关联规则
邻接矩阵
频繁模式树
data mining
association rules
adjacency matrix
frequent pattern tree
