Multivariate outlier detection in Stata
Vincenzo Verardi
University of Namur
(Centre for Research in the Economics of Development)
Namur, Belgium
and Université Libre de Bruxelles
(European Center for Advanced Research in Economics and Statistics
and Center for Knowledge Economics)
Brussels, Belgium
[email protected]
|
Catherine Dehon
Université libre de Bruxelles
(European Center for Advanced Research in Economics and Statistics
and Center for Knowledge Economics)
Brussels, Belgium
[email protected]
|
Abstract. Before implementing any multivariate statistical analysis based on empirical
covariance matrices, it is important to check whether outliers are present
because their existence could induce significant biases. In this article, we present
the minimum covariance determinant estimator, which is commonly used in robust
statistics to estimate location parameters and multivariate scales. These
estimators can be used to robustify Mahalanobis distances and to identify outliers.
Verardi and Croux (1999, Stata Journal 9: 439–453; 2010, Stata Journal
10: 313) programmed this estimator in Stata and made it available with the mcd
command. The implemented algorithm is relatively fast and, as we show in the
simulation example section, outperforms the methods already available in Stata,
such as the Hadi method.
View all articles by these authors:
Vincenzo Verardi, Catherine Dehon
View all articles with these keywords:
mcd, detection, multivariate outliers, robustness, minimum covariance determinant
Download citation: BibTeX RIS
Download citation and abstract: BibTeX RIS
|
