Communications for Statistical Applications and Methods

The Korean Statistical Society (한국통계학회)
권호 표지

A Study on High Breakdown Discriminant Analysis : A Monte Carlo Simulation

  • Moon Sup (Professor Department of Statistics Seoul National University) ;
  • Young Joo (Ph.D. Candidate, Department of Statistics, Seoul National University) ;
  • Youngjo (Associate Professor, Department of Statistics, Seoul National University)
  • Published : 2000.04.01
Download PDF
⟨ Previous Next ⟩

Abstract

The linear and quadratic discrimination functions based on normal theory are widely used to classify an observation to one of predefined groups. But the discriminant functions are sensitive to outliers. A high breakdown procedure to estimate location and scatter of multivariate data is the minimum volume ellipsoid or MVE estimator To obtain high breakdown classifiers outliers in multivariate data are detected by using the robust Mahalanobis distance based on MVE estimators and the weighted estimators are inserted in the functions for classification. A samll-sample MOnte Carlo study shows that the high breakdown robust procedures perform better than the classical classifiers.

Keywords

References

  1. Applied Statistics v.29 Robust Procedures in Multivariate Analysis Ⅰ: Robust covariance Estimation Campbell, N.A.
  2. Applied Statistics v.31 Robust Procedures in Multivariate Analysis Ⅱ: Robust Canonical Multivariate Analysis Campbell, N.A.
  3. Journal of Geochemistry Exploration v.43 Integrating a High-Breakdown Option into Discriminant Analysis in Exploration Geochemistry Chork, C.Y.;Rousseeuw, P.J.
  4. Journal of the American Statistical Association v.92 High-Breakdown Linear Discriminant Analysis Hawkins, D.M.;McLachlan, G.J.
  5. Applied Multivariate Statistical Analysis(3rd ed.) Johnson, R.A.;Wichern D.W.
  6. Journal of the American Statistical Association v.73 Discriminant Analysis Based on Rank Randles, R.H.;Broffitt, J.D.;Ramberg, J.S.;Hogg, R.V.
  7. Journal of the American Statistical Association v.73 Genetalized Linear and Quadratic Discriminant Functions Using Robust Estimates Randles, R.H.;Broffitt, J.D.;Ramberg, J.S.;Hogg, R.V.
  8. Robust Regression and Outlier Detection Rousseeuw, P.J.;Leroy, A.M.
  9. Journal of the American Statistical Association v.85 Unmasking Multivariate Outliers and Leverage Points Rousseeuw, P.J.;Zomeren, B.C.
  10. S-PLUS for Windows User's Manual Statistical Science
  11. Computational Statistics and Data Analysis v.17 Robust Two-Group Discrimination by Bounded Influence Regression. A Monte Carlo Simulation. Todorov, V.;Neykov, N.;Keytchev, P.