The Largest Nonidentiiable Outlier: a Comparison of Multivariate Simultaneous Outlier Identiication Rules

The aim of detecting outliers in a multivariate sample can be pursued in diierent ways. We investigate here the performance of several simultaneous multivariate outlier identiication rules based on robust estimators of location and scale. It has been shown that the use of estimators with high nite-sample breakdown point in such procedures yields a good behaviour with respect to the prevention of breakdown investigate by simulation, at which distance from the center of an underlying model distribution outliers can be placed until certain simultaneous identiication rules will detect them as outliers. We consider identiication procedures based on the minimum volume ellipsoid, the minimum covariance determinant, and S-estimators.