基于高阶项目集的频繁项目集发现算法

Title Frequent Itemsets Discovery Algorithm Base on the High-dimensional Itemsets

为了提高关联规则数据挖掘的效率,在研究Apriori算法原理和相关文献的基础上,提出了一种基于高阶项目集的频繁项目集发现算法.本算法不同于逐层迭代的搜索方式,而是采用从求解所有的高阶频繁m-项目集入手的方式,来发现隐藏在事务数据库中的频繁项目集.本算法避免了大量的候选项目集的产生,并且对数据库仅需进行有限次数的扫描,从而体现了算法的高效性.

In order to improve the efficiency of mining association rules,and based on the research on the Apriori algorithm and many related documents,a frequent itemsets discovery algorithm based on the high-dimensional itemsets is presented.Defferent from the method of iterative searching layer by layer,in order to discovering the frequent itemsets in the transactions database,the algorithm starts with solving all the high-dimensional frequent mitemsets.The algorithm avoids generating a mass of candidate itemsets,and the database is scanned only several times.So the efficiency of algorithm is very high.