一种改进的Apriori算法  

An improved Apriori algorithm

作　　者：赵军民[1] 张向娟[1] 高蔚[1] 

机构地区：[1]河南城建学院,河南平顶山467036

出　　处：《河南城建学院学报》2010年第6期60-62,共3页Journal of Henan University of Urban Construction

摘　　要：Apriori算法是挖掘关联规则频繁项集的最有影响的算法之一,它通过连接、剪枝等步骤产生频繁项集,进而产生强关联规则。由于面临海量数据,因此将会产生大量的候选项集,尤其是候选2-项集,严重影响了挖掘的效率。提出了一种改进的算法,此算法不产生小项候选集而直接产生大项候选集,从而提高了算法的效率。Apriori algorithm is one of the most influential algorithm for mining association rules in a frequent itemset,by connecting,pruning and other steps to produce less in the case of candidate itemsets generated frequent itemsets,and then generate strong association rules.In the face of massive data,so it will produce a large number of candidate items,especially the candidate 2-itemsets,thus seriously affecting the efficiency of mining.An improved algorithm,the algorithm does not produce the lesser candidate sets but large items directly from the candidate set to improve the efficiency of the algorithm.

关 键 词：数据挖掘 关联规则 APRIORI算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]