OUTLIER DETECTION IN HIGH DIMENSION BASED ON PROJECTION

摘要：

Outlier detection is one of the branches of data mining,with important applications in the domains of finance fraud detection, network intrusion analysis and so on. But most applications are high dimensional domains. Many algorithms use the concept of proximity to find outliers based on the relationship to the data set. However, the sparsity of high dimensional points results to the algorithms are not available for high dimensional space. In this paper, we discuss a new technique ODHDP(Outlier Detection in High Dimension based on Projection) which finds the outliers based on projection from the data set.

关键词： Data mining outlier high dimension projection

作者: PING GUO JI-YONG DAI YAN-XIA WANG

作者单位: School of Computer Science, Chongqing University, Chongqing, 400044, China

会议类型: 国际会议

会议名称: 2006 International Conference on Machine Learning and Cybernetics(IEEE第五届机器学习与控制论坛)

会议地点: 大连

会议语种:英文

页码: 1165-1169

在线出版日期: 2006-08-13（万方平台首次上网日期，不代表论文的发表时间）