机构地区: 云南大学数学与统计学院统计系
出 处: 《云南大学学报(自然科学版)》 1996年第4期383-387,共5页
摘 要: 关于正态样本下多个异常值的检验,各种文献中仅对异常值个数已知的情况作了讨论.当异常值个数未知时,本文成功的找到了确定异常值个数的方法.此法简单有效。 For the test of many outliers in normal samples,most literature only dealt with the case in which the number of outliers is known. When the k is unknown,in the paper,we successfully find a way of determing k,which can completely eliminate masking and swamping effect.