Advertisement

Computational Statistics

, Volume 27, Issue 3, pp 393–410 | Cite as

A comparison of algorithms for the multivariate L 1-median

  • Heinrich FritzEmail author
  • Peter Filzmoser
  • Christophe Croux
Original Paper

Abstract

The L 1-median is a robust estimator of multivariate location with good statistical properties. Several algorithms for computing the L 1-median are available. Problem specific algorithms can be used, but also general optimization routines. The aim is to compare different algorithms with respect to their precision and runtime. This is possible because all considered algorithms have been implemented in a standardized manner in the open source environment R. In most situations, the algorithm based on the optimization routine NLM (non-linear minimization) clearly outperforms other approaches. Its low computation time makes applications for large and high-dimensional data feasible.

Keywords

Algorithm Multivariate median Optimization Robustness 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Bedall F, Zimmermann H (1979) As 143: the mediancentre. Appl Stat 28: 325–328zbMATHCrossRefGoogle Scholar
  2. Chaudhuri P (1996) On a geometric notion of quantiles for multivariate data. J Am Statist Assoc 91: 862–872zbMATHCrossRefGoogle Scholar
  3. Croux C, Ruiz-Gazen A (2005) High breakdown estimators for principal components: the projection- pursuit approach revisited. J Multivar Anal 95: 206–226MathSciNetzbMATHCrossRefGoogle Scholar
  4. Croux C, Filzmoser P, Oliveira M (2007) Algorithms for projection-pursuit robust principal component analysis. Chemometr Intell Lab Syst 87: 218–225CrossRefGoogle Scholar
  5. Debruyne M, Hubert M, Van Horebeek J (2010) Detecting influential observations in Kernel PCA. Comput Stat Data Anal 54(12): 3007–3019MathSciNetCrossRefGoogle Scholar
  6. Dennis J, Schnabel R (1983) Numerical methods for unconstrained optimization and nonlinear equations. Prentice Hall, New JerseyzbMATHGoogle Scholar
  7. Fletcher R, Reeves C (1964) Function minimization by conjugate gradients. Comput J 7: 148–154MathSciNetCrossRefGoogle Scholar
  8. Gervini D (2008) Robust functional estimation using the median and spherical principal components. Biometrika 95: 587–600MathSciNetzbMATHCrossRefGoogle Scholar
  9. Gower J (1974) As 78: the mediancentre. J R Statist Soc 23: 466–470Google Scholar
  10. Hössjer O, Croux C (1995) Generalizing univariate signed rank statistics for testing and estimating a multivariate location parameter. Nonparametr Statist 4: 293–308MathSciNetzbMATHCrossRefGoogle Scholar
  11. Lopuhaä H, Rousseeuw P (1991) Breakdown points of affine equivariant estimators of multivariate location and covariance matrices. Ann Stat 19(1): 229–248zbMATHCrossRefGoogle Scholar
  12. Nelder J, Mead R (1965) A simplex algorithm for function minimization. Comput J 7: 308–313zbMATHGoogle Scholar
  13. Nocedal J, Wright S (2006) Numerical optimization, 2nd edn. Springer, New YorkzbMATHGoogle Scholar
  14. R Development Core Team (2010) R: a Language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org, ISBN 3-900051-07-0
  15. Reimann C, Filzmoser P, Garrett R, Dutter R (2008) Statistical data analysis explained. Applied environmental statistics with R. Wiley, ChichesterCrossRefGoogle Scholar
  16. Schnabel R, Koontz J, Weiss B (1985) A modular system of algorithms for unconstrained minimization. ACM Trans Math Softw 11: 419–440MathSciNetzbMATHCrossRefGoogle Scholar
  17. Serneels S, Croux C, Filzmoser P, Van Espen P (2005) Partial robust M-regression. Chemometr Intell Lab Syst 79: 55–64CrossRefGoogle Scholar
  18. Vardi Y, Zhang CH (2000) The multivariate l 1-median and associated data depth. Proc Natl Acad Sci USA 97(4): 1423–1426MathSciNetzbMATHCrossRefGoogle Scholar
  19. Weber A (1909) Über den Standort der Industrien. Mohr, TübingenGoogle Scholar
  20. Weiszfeld E (1937) Sur le point pour lequel la somme des distances de n points donnés est minimum. Tôhoku Math J 43: 355–386Google Scholar

Copyright information

© Springer-Verlag 2011

Authors and Affiliations

  • Heinrich Fritz
    • 1
    Email author
  • Peter Filzmoser
    • 1
  • Christophe Croux
    • 2
    • 3
  1. 1.Department of Statistics and Probability TheoryVienna University of TechnologyViennaAustria
  2. 2.Faculty of Business and EconomicsK. U. Leuven UniversityLeuvenBelgium
  3. 3.Faculty of Business and EconomicsTilburg UniversityTilburgThe Netherlands

Personalised recommendations