摘要
多元数据中的Outliers的不一致性检验与一元数据的Outliers的不一致性检验同样重要,而且是相关联的.本文讨论了m元正态样本的n个个体的马氏距离的极端值,并给出检验这些极端值不一致性的方法.
The tests of discordancy of outliers have as much relevance and importance for multivariate data as they do for univariat data. In this paper the extreme value of the Mahalanobis distances of n individual points in the m-variat normal sample is discussed and the method of tests for discordancy of outliers is also given.
出处
《云南大学学报（自然科学版）》
CAS
CSCD
1990年第1期10-18,共9页
Journal of Yunnan University(Natural Sciences Edition)
关键词
多元正态
OUTLIERS
不一致性检验
Outliers, Mahalanobis distance, tests, multivariate normal distribution.
