A Comparative Study of Outlier Detection Procedures in Multiple Linear Regression

被引：0

作者：

Ampanthong, Pimpan ^{[1
]}

Suwattee, Prachoom ^{[1
]}

机构：

[1] Natl Inst Dev Adm, Sch Appl Stat, Bangkok, Thailand

来源：

IMECS 2009: INTERNATIONAL MULTI-CONFERENCE OF ENGINEERS AND COMPUTER SCIENTISTS, VOLS I AND II | 2009年

关键词：

Multiple linear regression; Outliers; Outlier detection; Residuals; MULTIVARIATE LOCATION; HIGH-BREAKDOWN; ROBUST;

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Outlier detection methods in multiple linear regression are reviewed. Eight statistics for outlier detection have been investigated and compared. It is found from Monte Carlo simulation that Mahalanobis distance (MDi) identifiers the presence of outliers more often than the others for small, medium and large sample sizes with different percentages outliers in the regressors and in both the regressors; and the dependent variable. The next best statistics for the detection are Hat matrix (h(ii)) Cook's square distance (CDi) and DEFFITi distance. As for the dependent variable outlier, Cook's square distance (CDi) and PRESS residual (r((i))) perform better than the others.

引用

页码：704 / 709

页数：6

共 27 条

[1]

Atkinson Anthony Curtes, 1985, Plots, transformations and regression

[2]

an introduction to graphical methods of diagnostic regression analysis

[3]

Barnett V., 1994, Outliers in statistical data

[4]

Birkes D., 1993, Alternative Methods of Regression

[5]

Campbell N. A., 1980, Applied Statistics, V29, P231, DOI 10.2307/2346896

[6]

HADI AS, 1994, J ROY STAT SOC B MET, V56, P393

[7]

HADI AS, 1992, J ROY STAT SOC B MET, V54, P761

[8] PROCEDURES FOR THE IDENTIFICATION OF MULTIPLE OUTLIERS IN LINEAR-MODELS [J].

HADI, AS ;

SIMONOFF, JS .

JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1993, 88 (424) :1264-1272

[9]

Huber Peter, 1981, International Encyclopedia of Statistical Science

[10] A MONTE-CARLO COMPARISON OF 5 PROCEDURES FOR IDENTIFYING OUTLIERS IN LINEAR-REGRESSION [J].

KIANIFARD, F ;

SWALLOW, WH .

COMMUNICATIONS IN STATISTICS-THEORY AND METHODS, 1990, 19 (05) :1913-1938

← 1 2 3 →