稳健回归分析  

Robust regression analysis

作　　者：滕素珍[1] 

机构地区：[1]大连理工大学应用数学系

出　　处：《大连理工大学学报》1991年第6期719-723,共5页Journal of Dalian University of Technology

摘　　要：The article suggests precautionary measures which should be taken when outliers appear in experimental or observational data That is to establish robust regression with an insensitive statistical method. A relationship between weight and trimmed constant is derived. The reason is stated for the purpose of forming additional equations. Explanatory notes are given to the iteration formula. Robust regression is applied to forecast. It is proved that robust regression has advantages over the classical regression.The article suggests precautionary measures which should be taken when outliers appear in experimental or observational data That is to establish robust regression with an insensitive statistical method. A relationship between weight and trimmed constant is derived. The reason is stated for the purpose of forming additional equations. Explanatory notes are given to the iteration formula. Robust regression is applied to forecast. It is proved that robust regression has advantages over the classical regression.

关 键 词：回归分析 线性模型 M-估计 稳健 

分 类 号：O212.1[理学—概率论与数理统计]