摘要
文章简要介绍了关联规则挖掘技术,通过对经典算法FP-Growth在某鞋业ERP系统中的应用分析,发现了FP-Growth算法挖掘的部分数据无用的缺陷,提出了一种增加协方差兴趣度阈值的改进算法—CovFP-Growth算法.该算法采用协方差的概念,能够更准确地挖掘出交易集中不同产品间的紧密相关性,减少产生的无用规则,为ERP的产品预测计划决策提供了良好的理论依据和实现方法.
This paper introduces association rules mining technology briefly. By analyzing the application of classic FP-Growth algorithm in ERP System, and finds out FP-Growth algorithm＇s limitation which often generates some useless datas. So this paper gives a CovFP- Growth algorithm Based covariance Measure for Updating FP-Growth. Using the covariance method, CovFP-Growth can mine the close relations among different products in the database, decrease the amount of useless rules and provide effective assistant method for the generating of prediction plan.
出处
《数学的实践与认识》
CSCD
北大核心
2008年第12期11-18,共8页
Mathematics in Practice and Theory
基金
四川省宜宾市科技局科研基金(200702036)
宜宾学院青年基金(QJ05-08)
