The detection of multipleoutliers in random effects models of the unbalanced one-way classifications is studied in this paper.
研究了非平衡单向分类随机效应模型中多个异常值的检验问题。
2
Objective:this research is to deal with the masking effect when multipleoutliers occur during the traditional diagnostic regression analysis.
目的:本研究试图解决多个异常点情况下经典回归诊断方法中存在的掩盖现象。
3
When multipleoutliers occur in linear regression model or the distribution of residuals is not normal, we can use residuals rank as weight function to get some resist estimator.