基于SQL分层抽样的数据挖掘算法的改进  

Empirical study of data mining algorithm based on SQL stratified sampling

作　　者：谢笑盈[1] 邢君飞[1] 

机构地区：[1]浙江师范大学数理与信息工程学院,浙江金华321004

出　　处：《浙江师范大学学报（自然科学版）》2011年第2期175-178,共4页Journal of Zhejiang Normal University:Natural Sciences

摘　　要：在对数据库聚类分析的基础上进行分层抽样,并使用关联规则,得出了数据之间的潜在关系.同时,对网民健身情况调查数据进行了实证分析,在SQL Server 2005上实现了抽样后的关联规则挖掘,提高了关联的效率,并取得了良好的效果.另外,对关联规则的评估作了一定的改进创新.It was aimed to get out the potential relation by the association rule algorithm in data mining after the stratified sampling based on clustering algorithms carried out in database.Meanwhile,an empirical study about netizen keeping fit condition on association rule data mining based on sampling was realized in SQL Server 2005,optimizing the efficiency of association,and obtained good effect.On the other hand,a new evaluation rule about association was also attempted.

关 键 词：聚类分析 分层抽样 关联规则 SQL 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]