Confirmation of multiple outliers in generalized linear and nonlinear regressions

被引：13

作者：

Lee, AH

Fung, WK

机构：

[1] NO TERR UNIV, FAC SCI, DARWIN, NT 0909, AUSTRALIA

[2] UNIV HONG KONG, DEPT STAT, HONG KONG, HONG KONG

来源：

COMPUTATIONAL STATISTICS & DATA ANALYSIS | 1997年 / 25卷 / 01期

关键词：

generalized linear model; high breakdown methods; multiple outliers; nonlinear regression; simulated envelopes;

D O I：

10.1016/S0167-9473(96)00086-2

中图分类号：

TP39 [计算机的应用];

学科分类号：

081203 ; 0835 ;

摘要：

We propose a stepwise procedure for the detection of multiple outliers in generalized linear models and nonlinear regressions. The algorithm starts with a high breakdown point estimation method to find the potential outliers, then uses an adding-back iterative approach to confirm such outliers with envelopes derived from simulation. The proposed method can overcome the masking and swamping problems commonly occurred in multiple outliers detection. We apply the procedure to two real examples, and satisfactory results are obtained. (C) 1997 Elsevier Science B.V.

引用

页码：55 / 65

页数：11

共 28 条

[1]

[Anonymous], 1989, STAT ANAL DISCRETE D

[2]

BARNETT V, 1994, OUTLIERS STATISTICAL

[3]

CARROLL RJ, 1993, J ROY STAT SOC B MET, V55, P693

[4]

CHRISTMANN A, 1994, BIOMETRIKA, V81, P413, DOI 10.2307/2336973