基于改进的K-means算法的关联规则数据挖掘研究

关联规则是数据挖掘中的概念,通过分析数据找到数据之间的关联.海量数据会产生大量冗余和相似的关联规则,影响用户对规则的理解和判断.本文采用鸢尾花数据集进行实验.建立三个检验指标,删除冗余关联规则;在进行K-means分析时利用规则产生的三角形迭代选择初始点,再将删除冗余后的规则进行聚类.实验证实本文方法将相似的关联规则归为一簇,能有效的帮助用户迅速找到有用的关联规则,有助于用户更好的对规则进行理解和分析,提高了聚类的效率.

Research on Association Rule Data Mining Based on Improved K-means Algorithm

Association rules are a concept in data mining.The association betw een data is found by analyzing the data.Massive data will generate a large number of redundant and similar association rules,affecting users’understanding and judgment of the rules.This article uses the iris data set for experiments.First,three test indexes are established to delete redundant association rules.When performing K-means analysis,the triangles generated by the rules are used to iteratively select the initial point,and then the redundant rules are clustered.The experiments confirm that the method in this paper classifies similar association rules into a cluster,w hich can effectively help users quickly find useful association rules,help users better understand and analyze the rules,and improve the efficiency of clustering.