Outliers robustness in multivariate orthogonal regression

Author

Calafiore, Giuseppe Carlo

Author_Institution

Dipartimento di Autom. e Inf., Politecnico di Torino, Italy

Volume

30

Issue

6

fYear

2000

fDate

11/1/2000 12:00:00 AM

Firstpage

674

Lastpage

679

Abstract

Deals with the problem of multivariate affine regression in the presence of outliers in the data. The method discussed is based on weighted orthogonal least squares. The weights associated with the data satisfy a suitable optimality criterion and are computed by a two-step algorithm requiring a RANSAC step and a gradient-based optimization step. Issues related to the breakdown point of the method are discussed, and examples of application on various real multidimensional data sets are reported in the paper

Keywords

estimation theory; least squares approximations; matrix algebra; statistical analysis; RANSAC step; breakdown point; gradient-based optimization step; multidimensional data sets; multivariate affine regression; multivariate orthogonal regression; optimality criterion; outliers robustness; two-step algorithm; weighted orthogonal least squares; Electric breakdown; Equations; Gaussian processes; Helium; History; Least squares methods; Multidimensional systems; Noise robustness; Pattern recognition; Regression analysis;

fLanguage

English

Journal_Title

Systems, Man and Cybernetics, Part A: Systems and Humans, IEEE Transactions on

Publisher

ieee

ISSN

1083-4427

Type

jour

DOI

10.1109/3468.895890

Filename

895890