摘要
文中研究基于兼存率(多个项同时存在的概率)与单项事务(仅包含一个项的事务)筛选提出关联规则优化算法ARO,通过对数据集D中每个项与事务T进行处理来过滤无用或干扰的数据,从而得出更加准确、显著的关联规则。实验结果表明,在标准数据集中,对比传统算法META,ARO算法在关联规则分析的显著性与准确性方面均有性能提升。
In this paper,an Association Rule Optimization Algorithm( ARO) is proposed based on MCP( Probability of multiple items coexisting) and TOI( a transaction has only one item). By eliminating useless or disruptive data with a selection strategy of dealing with each item in the datasets D and transaction T, more accutate and significant association rules are reached. Compared with META algorithm,the ARO algrithm has effectively improved the significance and accuracy of association rules.
作者
田榆杰
宋耀莲
龙华
张漪
TIAN Yu-jie;SONG Yao-lian;LONG Hua;ZHANG Yi(Kunming University of Science and Technology,Faculty of Information Engineering and Automation,Kunming 650000,China)
出处
《信息技术》
2019年第1期75-78,共4页
Information Technology
