摘要
针对Apriori算法的不足之处,提出了基于矩阵的算法,该算法首先将事务数据库用矩阵表示出来,并对矩阵进行处理,找出包含最多项的频繁K-项集,最后再利用矩阵找出从频繁2-项集到频繁K-1项集的所有频繁项集。通过一个实例表明了该算法的具体实现过程,并与其它算法进行比较,阐述了该算法的优缺点。该算法不但充分利用了矩阵这一工具,用"与运算"的方法代替了到数据库中去查找的算法,而且大大减少了候选频繁项集的产生,从而节省了计算频繁项集的时间,提高了计算的效率。
To the deficiency of Apriori algorithm, an improved Apriori algorithm based on the matrix is put forward. This algorithm converts the affair database to a matrix and operates it to find out the largest K-frequent itemsets, at last uses the matrix to find out all the other frequent itemsets from 2-frequent itemsets to K-1-frequent itemsets. Using an example shows the concrete process of the algorithm and comparing with other algorithms the advantages and disadvantages of the algorithm are expounded. This algorithm not only uses matrix and ＂AND operation＂ instead of the query operation in the database, but also greatly reduce the number of candidates of frequent itemsets, so the algorithm obtains the bonus time of calculating and improves the efficiency of computing.
出处
《计算机工程与设计》
CSCD
北大核心
2009年第10期2435-2438,共4页
Computer Engineering and Design
基金
国家863高技术研究发展计划基金项目(2007AA01Z185)
关键词
矩阵
与运算
频繁项集
最小支持度
事务
Matrix
AND operation
frequent itemsets
Minsupport
affair
