Acta Mathematicae Applicatae Sinica

, Volume 21, Issue 2, pp 209–224

A Review of Robust Regression and Diagnostic Procedures in Linear Regression

Original Papers

DOI: 10.1007/s10255-005-0230-2

Cite this article as:
Jajo, N.K. Acta Mathematicae Applicatae Sinica, English Series (2005) 21: 209. doi:10.1007/s10255-005-0230-2

Abstract

The concern over outliers is old since Bernoulli (see [12]), reviewed historically by [11] and updated with [10] in their encyclopedia textbook. James et al.[46] used simulation technique to compare some recent published outlier detection procedures.

The history of adept and diagnosis of outliers is traced from old and presence comments. Theil-type or Rank, Brown-Mood, Lp, M, adaptive M, GM, and Trimmed-Winsorization estimators are the most popular estimators that we will review in this paper as an application to outlier accommodation. We will review and compare the most numerical and graphical displays based on residuals to flag outliers.

Keywords

Masking outliers robust estimator swamping 

2000 MR Subject Classification

62G35 

Copyright information

© Springer-Verlag Berlin Heidelberg 2005

Authors and Affiliations

  1. 1.School of Quantitative Methods & Mathematical Sciences, College of Law and BusinessUniversity of Western Sydney, Blacktown CampusPenrith South DC NSW 1797Australia

Personalised recommendations