摘要
现存有关孤立点分析的研究很少解释识别出的孤立点的用户意义,而孤立点通常都包含着重要的信息,在许多应用领域中对于孤立点意义的解释和孤立点本身同等重要。因此,给出孤立点用户意义的定义,并提出一种基于距离和的孤立点用户意义分析算法(DSCM),对每一孤立点给出相应的解释,以帮助用户更好地理解孤立数据。应用到质量管理中的结果表明,该算法是有效的和实用的,且易用性较强。
The customer meaning for outlier explanation is rarely provided in the current studies. The outliers usually contain important information, and for many applications, the explanations are as important to the user as the outliers. A new definition of outlier customer meaning was given, and a new outlier customer meaning analysis algorithm named DSCM was put forward based on distance sum. The algorithm gave an explanation of every outlier, which improved the user's understanding of the data. Then the algorithm was applied to quality management, and the results show that the algorithm is effective and practical, and more easy to use.
出处
《计算机应用》
CSCD
北大核心
2009年第11期3077-3079,共3页
journal of Computer Applications
基金
重庆市科技攻关资金资助项目(CSTC
2009AB2049
CSTC
2009AC2068)
关键词
孤立点
用户意义
距离和
质量管理
数据挖掘
outlier
customer meaning
distance sum
quality management
data mining