Constrained \(M\)-estimation for multivariate location and scatter. (English) Zbl 0862.62048

Summary: Consider the problem of estimating the location vector and scatter matrix from a set of multivariate data. Two standard classes of robust estimates are \(M\)-estimates and \(S\)-estimates. The \(M\)-estimates can be tuned to give good local robustness properties, such as good efficiency and a good bound on the influence function at an underlying distribution such as the multivariate normal. However, \(M\)-estimates suffer from poor break-down properties in high dimensions. On the other hand, \(S\)-estimates can be tuned to have good breakdown properties, but when tuned in this way, they tend to suffer from poor local robustness properties.
In this paper a hybrid estimate called a constrained \(M\)-estimate is proposed which combines both good local and good global robustness properties.


62H12 Estimation in multivariate analysis
62F35 Robustness and adaptive procedures (parametric inference)
Full Text: DOI


[1] DAVIES, P. L. 1987. Asy mptotic behavior of S-estimates of multivariate location parameters and dispersion matrices. Ann. Statist. 15 1269 1292. Z. · Zbl 0645.62057 · doi:10.1214/aos/1176350505
[2] DONOHO, D. L. 1982. Breakdown properties of multivariate location estimates. Ph.D. qualifying paper, Dept. Statistics, Harvard Univ. Z.
[3] DONOHO, D. L. and HUBER, P. J. 1983. The notion of breakdown point. In A Festschrift for Z. Erich L. Lehmann P. J. Bickel, K. A. Doksum and J. L. Hodges, eds. 157 184. Wadsworth, Belmont, CA. Z. · Zbl 0523.62032
[4] HAMPEL, F. R., RONCHETTI, E. M., ROUSSEEUW, P. J. and STAHEL, W. A. 1986. Robust Statistics: The Approach Based on Influence Functions. Wiley, New York. · Zbl 0593.62027
[5] HUBER, P. J. 1977. Robust covariances. In Statistical Decision Theory and Related Topics S. S.. Guta and D. S. Moore, eds. 165 191. Academic Press, New York. Z. · Zbl 0435.62041
[6] HUBER, P. J. 1981. Robust Statistics. Wiley, New York. Z. · Zbl 0536.62025
[7] KELKER, D. 1970. Distribution theory of spherical distributions and a location scale parameter generalization. Sankhy a Ser. A 32 419 430. Z. · Zbl 0223.60008
[8] KENT, J. T. and Ty LER, D. E. 1991. Redescending M-estimates of multivariate location and scatter. Ann. Statist. 19 2102 2119. Z. · Zbl 0763.62030 · doi:10.1214/aos/1176348388
[9] KIEFER, J. and WOLFOWITZ, J. 1956. Consistency of the maximum likelihood estimator in the presence of infinitely many incidental parameters. Ann. Math. Statist. 27 887 906. Z. · Zbl 0073.14701 · doi:10.1214/aoms/1177728066
[10] LOPUHAA, H. P. 1989. On the relationship between S-estimators and M-estimators of multi\" variate location and covariance. Ann. Statist. 17 1662 1683. Z. · Zbl 0702.62031 · doi:10.1214/aos/1176347386
[11] MARONNA, R. A. 1976. Robust M-estimates of multivariate location and scatter. Ann. Statist. 4 51 67.Z. · Zbl 0322.62054 · doi:10.1214/aos/1176343347
[12] MUIRHEAD, R. J. 1982. Aspects of Multivariate Statistical Theory. Wiley, New York. Z. · Zbl 0556.62028
[13] ROUSSEEUW, P. J. 1985. Multivariate estimation with high breakdown point. In Mathematical Z. Statistics and Applications W. Grossmann, G. Pflug, I. Vincze and W. Wertz, eds. 283 297. Reidel, Dordrecht. Z. · Zbl 0609.62054
[14] STAHEL, W. A. 1981. Breakdown of covariance estimators. Technical Report 31, Pachgruppe fur \" Statistik, ETH, Zurich. Z.
[15] Ty LER, D. E. 1983. Robustness and efficiency properties of scatter matrices. Biometrika 70 411 420. Z. JSTOR: · Zbl 0536.62042 · doi:10.1093/biomet/70.2.411
[16] Ty LER, D. E. 1991. Some issues in the robust estimation of multivariate location and scatter. In Z. Directions in Robust Statistics and Diagnostics 2 W. Stahel and S. Weisberg, eds. 327 336. Springer, New York.
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.