Some Extensions of Tukey's Depth Function

Jian Zhang

Journal of Multivariate Analysis, 2002, vol. 82, issue 1, 134-165

Abstract: As the extensions of Tukey's depth, a family of affine invariant depth functions are introduced for multivariate location and dispersion. The location depth functions can be used for the purpose of multivariate ordering. Such kind ordering can retain more information from the original data than that based on Tukey's depth. The dispersion depth functions provide some additional view of the dispersion of the data set. It is shown that these sample depth functions converge to their population versions uniformly on any compact subset of the parameter space. The deepest points of these depth functions are affine equivariant estimates of multivariate location and dispersion. Under some general conditions these estimates are proved to have asymptotic breakdown points at least 1/3 and convergence rates of 1/. Their asymptotic distributions are also obtained under some regularity conditions. A new algorithm based on the idea of thresholding is presented for computing these kinds of estimates and realized in the bivariate case. Simulations indicate that some of them could have the empirical mean squared errors smaller than those based on Tukey's depth function or Donoho's depth function.

Keywords: depth; and; outlyingness; location; and; dispersion; estimates; breakdown; point; multivariate; ordering (search for similar items in EconPapers)
Date: 2002
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (4)

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0047-259X(01)92011-1
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:jmvana:v:82:y:2002:i:1:p:134-165

Ordering information: This journal article can be ordered from
http://www.elsevier.com/wps/find/supportfaq.cws_home/regional
https://shop.elsevie ... _01_ooc_1&version=01

Access Statistics for this article

Journal of Multivariate Analysis is currently edited by de Leeuw, J.

More articles in Journal of Multivariate Analysis from Elsevier
Bibliographic data for series maintained by Catherine Liu ().