Robustness Properties ofkMeans and TrimmedkMeans
Authors:
- 1 September 1999
- journal article
- research article
- Published by Taylor & Francis in Journal of the American Statistical Association
- Vol. 94 (447) , 956-969
Abstract
The generalized k means method is based on the minimization of the discrepancy between a random variable (or a sample of this random variable) and a set with k points measured through a penalty function Φ. As in the M estimators setting (k = 1), a penalty function, Φ, with unbounded derivative, Ψ, naturally leads to nonrobust generalized k means. However, surprisingly the lack of robustness extends also to the case of bounded Ψ; that is, generalized k means do not inherit the robustness properties of the M estimator from which they came. Attempting to robustify the generalized k means method, the generalized trimmed k means method arises from combining k means idea with a so-called impartial trimming procedure. In this article study generalized k means and generalized trimmed k means performance from the viewpoint of Hampel's robustness criteria; that is, we investigate the influence function, breakdown point, and qualitative robustness, confirming the superiority provided by the trimming. We include the study of two real datasets to make clear the robustness of generalized trimmed k means.Keywords
This publication has 21 references indexed in Scilit:
- A central limit theorem for multivariate generalized trimmed $k$-meansThe Annals of Statistics, 1999
- Principal Points and Self-Consistent Points of Elliptical DistributionsThe Annals of Statistics, 1995
- On the breakdown point of multivariate location estimators based on trimming proceduresStatistics & Probability Letters, 1991
- A Central Limit Theorem for $k$-Means ClusteringThe Annals of Probability, 1982
- Nonparametric Interval and Point Prediction Using Data Trimmed by a Grubbs-Type Outlier RuleThe Annals of Statistics, 1982
- Strong Law of Large Numbers for Measures of Central Tendency and Dispersion of Random Variables in Compact Metric SpacesThe Annals of Statistics, 1981
- Asymptotic Distributions for Clustering CriteriaThe Annals of Statistics, 1978
- Generalized Means and Associated Families of DistributionsThe Annals of Mathematical Statistics, 1969
- On Grouping for Maximum HomogeneityJournal of the American Statistical Association, 1958
- Note on GroupingJournal of the American Statistical Association, 1957