基于维间扩展和事务压缩的关联规则算法改进  

Improvement of associative rule algorithm based on dimensional expansion and transaction reduction

作　　者：张云洋[1] 刘芳[2] 

机构地区：[1]西藏大学图书馆,西藏拉萨850012 [2]西藏大学工学院计算机系

出　　处：《计算机时代》2012年第9期24-26,30,共4页Computer Era

基　　金：西藏大学青年科研培育基金项目"基于数据挖掘的藏文网页搜索算法研究"(ZDPJZK201202)

摘　　要：Apriori是关联规则挖掘的经典算法,在利用该算法进行医疗数据挖掘的过程中,发现其频繁项集产生过程有一些不必要的开销,为此提出了改进算法Mypriori,利用维间扩展和事务压缩策略来提高频集发现的效率,并通过实验验证了算法的有效性。Apriori algorithm is a classic algorithm for associative rules. When applying to medical data mining, Apriori will generate some unnecessary operations in the process of finding frequent item sets. An improved algorithm named Mypriori is presented. Mypriori uses two tactics to enhance frequent item sets producing, which are called dimensional expansion and transaction reduction. Mypriori is proved to be effective through data mining experiments.

关 键 词：关联规则 维间扩展 事务压缩 APRIORI Mypriori 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]