摘要
讨论了均值漂移模型与数据删除模型在异常点的检验上的等价性作用.同时阐述了两个具有某种相关关系的异常点对回归系数的影响,并把它推广到多个异常点上,由此得到几个重要结论.
In this paper, the equivalence role of mean shift outlier model and case deletion model in the testing anomalies are discussed. At the same time, the influence of the two with some correlation relation between the anomalous points on the regression coefficients is introduced, and it is extended to multiple outliers. Thus some important conclusions are obtained.
出处
《哈尔滨师范大学自然科学学报》
CAS
2013年第3期32-35,共4页
Natural Science Journal of Harbin Normal University
关键词
异常点
最小二乘估计
均值漂移模型
数据删除模型
Outlier
Least square estimation
Mean - shift outlier model
Case deletion model