Computation of robust estimates of multivariate location and shape

David M Rocke, D. L. Woodruff

Research output: Contribution to journalArticle

37 Citations (Scopus)

Abstract

This paper reviews the state of the art in the computation of robust estimates of multivariate location and shape using combinatorial estimators such as the minimum volume ellipsoid (MVE) and iterative M‐ and S‐estimators. We also present new results on the behavior of M‐ and S‐estimators in the presence of different types of outliers, and give the first computational evidence on compound estimators that use the MVE as a starting point for an S‐estimator. Problems with too many data points in too many dimensions cannot be handled by any available technology; however, the methods presented in this paper substantially extend the size of problem that can be successfully handled.

Original languageEnglish (US)
Pages (from-to)27-42
Number of pages16
JournalStatistica Neerlandica
Volume47
Issue number1
DOIs
StatePublished - 1993

Fingerprint

Minimum Volume Ellipsoid
Robust Estimate
Estimator
Outlier

Keywords

  • breakdown
  • genetic algorithms
  • heuristic search
  • minimum volume ellipsoid
  • M‐estimators
  • outliers
  • simulated anneating
  • tabu search

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Computation of robust estimates of multivariate location and shape. / Rocke, David M; Woodruff, D. L.

In: Statistica Neerlandica, Vol. 47, No. 1, 1993, p. 27-42.

Research output: Contribution to journalArticle

@article{b967d01f08d84a0895d50d312848e4e1,
title = "Computation of robust estimates of multivariate location and shape",
abstract = "This paper reviews the state of the art in the computation of robust estimates of multivariate location and shape using combinatorial estimators such as the minimum volume ellipsoid (MVE) and iterative M‐ and S‐estimators. We also present new results on the behavior of M‐ and S‐estimators in the presence of different types of outliers, and give the first computational evidence on compound estimators that use the MVE as a starting point for an S‐estimator. Problems with too many data points in too many dimensions cannot be handled by any available technology; however, the methods presented in this paper substantially extend the size of problem that can be successfully handled.",
keywords = "breakdown, genetic algorithms, heuristic search, minimum volume ellipsoid, M‐estimators, outliers, simulated anneating, tabu search",
author = "Rocke, {David M} and Woodruff, {D. L.}",
year = "1993",
doi = "10.1111/j.1467-9574.1993.tb01404.x",
language = "English (US)",
volume = "47",
pages = "27--42",
journal = "Statistica Neerlandica",
issn = "0039-0402",
publisher = "Wiley-Blackwell",
number = "1",

}

TY - JOUR

T1 - Computation of robust estimates of multivariate location and shape

AU - Rocke, David M

AU - Woodruff, D. L.

PY - 1993

Y1 - 1993

N2 - This paper reviews the state of the art in the computation of robust estimates of multivariate location and shape using combinatorial estimators such as the minimum volume ellipsoid (MVE) and iterative M‐ and S‐estimators. We also present new results on the behavior of M‐ and S‐estimators in the presence of different types of outliers, and give the first computational evidence on compound estimators that use the MVE as a starting point for an S‐estimator. Problems with too many data points in too many dimensions cannot be handled by any available technology; however, the methods presented in this paper substantially extend the size of problem that can be successfully handled.

AB - This paper reviews the state of the art in the computation of robust estimates of multivariate location and shape using combinatorial estimators such as the minimum volume ellipsoid (MVE) and iterative M‐ and S‐estimators. We also present new results on the behavior of M‐ and S‐estimators in the presence of different types of outliers, and give the first computational evidence on compound estimators that use the MVE as a starting point for an S‐estimator. Problems with too many data points in too many dimensions cannot be handled by any available technology; however, the methods presented in this paper substantially extend the size of problem that can be successfully handled.

KW - breakdown

KW - genetic algorithms

KW - heuristic search

KW - minimum volume ellipsoid

KW - M‐estimators

KW - outliers

KW - simulated anneating

KW - tabu search

UR - http://www.scopus.com/inward/record.url?scp=84995007758&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84995007758&partnerID=8YFLogxK

U2 - 10.1111/j.1467-9574.1993.tb01404.x

DO - 10.1111/j.1467-9574.1993.tb01404.x

M3 - Article

AN - SCOPUS:84995007758

VL - 47

SP - 27

EP - 42

JO - Statistica Neerlandica

JF - Statistica Neerlandica

SN - 0039-0402

IS - 1

ER -