加权关联规则的改进算法  

An Improved Algorithm for Mining Association Rules With Weighted Items

作　　者：宫雨[1] 武森[1] 尹阿东[1] 高学东[1] 

机构地区：[1]北京科技大学管理学院,北京100083

出　　处：《计算机工程与应用》2004年第22期177-179,188,共4页Computer Engineering and Applications

摘　　要：论文讨论了加权关联规则问题,针对布尔类型的加权关联规则问题提出一种改进算法。该算法首先利用普通的关联规则算法产生频繁集,然后在该频繁集的基础上产生加权频繁集。同时,给出了最优的最小支持度设定方法,保证了普通关联规则算法所产生的频繁集为加权频繁集的超集。该算法有较高的效率,并且能够有效利用已有的关联规则算法。This paper discusses algorithms for mining association rules with weighted items.An improved algorithm for mining association rules with weighted items is provided.This algorithm first produces frequent items using previous common association rules mining algorithm,then produces weighted frequent items from the previous frequent items.This paper also discusses how to set the optimal minimum support for the common association rules mining algorithm,which can guarantee the frequent items are the weighted frequent items' superset.This algorithm is quite efficient ,and it can use existing association rules mining algorithms.

关 键 词：数据挖掘 关联规则 支持度 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]