期刊文献+

基于多数据库的模糊元关联规则挖掘方法 被引量:3

A METHOD OF MINING FUZZY META ASSOCIATION RULES FROM MULTIDATABASES
下载PDF
导出
摘要 分布式环境下,数据的特征是规模大、分布式存储。采用传统的数据挖掘技术分别对每个数据集进行分析一般都容易实现,但是要对全部数据进行整体决策时就比较困难。为此,提出一种新的挖掘方法,可以从多个数据集中挖掘规则。提出元关联规则生成模型,可以发现在每个独立的数据集中挖掘的规则之间的共同联系。设计清晰元关联规则和模糊元关联规则两种框架,对清晰元关联规则挖掘算法和模糊元关联规则挖掘算法做了对比。结果表明,模糊元关联规则挖掘方法在易用性和精确性方面比清晰方法要好。 In distributed environment,data is characterized by large scale and distributed storage. Generally,it is easy to achieve to use traditional data mining techniques to analyze each data set separately. However,it is more difficult to make an overall decision on all the data. Therefore,a new mining method is proposed in this paper,which can be used to mine rules from multiple data sets. A generation model is put forward based on meta association rule,finding out the relationship between the rules mined in each independent data set. Two kinds of frameworks,crisp meta association rules and fuzzy meta association rules,are designed. And comparison between their algorithms is also made. The results show that the mining method based on the fuzzy meta association rules is better than the crisp method in terms of ease of use and accuracy.
作者 刘小燕 王健 Liu Xiaoyan;Wang Jian(School of Computer Science and Technology, Henan Polytechnic University, Jiaozuo 454000, Henan, China;School of Safety Science and Engineering, Henan Polytechnic University, Jiaozuo 454000, Henan, China)
出处 《计算机应用与软件》 北大核心 2018年第5期42-47,123,共7页 Computer Applications and Software
基金 国家自然科学基金项目(11601129) 河南省矿山信息化实验室开放基金项目(KY2014-04)
关键词 关联规则 元关联规则 模糊 清晰 多数据库 Association rule Meta association rule Fuzzy Crisp Muhidatabases
  • 相关文献

参考文献5

二级参考文献65

共引文献17

同被引文献23

引证文献3

二级引证文献12

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部