ADVERTISEMENT

If you are seeing this message, you may be experiencing temporary network problems. Please wait a few minutes and refresh the page. If the problem persists, you may wish to report it to your local Network Manager.

It is also possible that your web browser is not configured or not able to display style sheets. In this case, although the visual presentation will be degraded, the site should continue to be functional. We recommend using the latest version of Microsoft or Mozilla web browser to help minimise these problems.

Wiley InterScience

Scandinavian Journal of Statistics

Scandinavian Journal of Statistics

Volume 33 Issue 4, Pages 753 - 763

Published Online: 17 Mar 2006

© 2009 Board of the Foundation of the Scandinavian Journal of Statistics



< Previous Abstract  |  Next Abstract >

Save Article to My Profile      Download Citation      Request Permissions

Abstract |  References  |  Full Text: HTML, PDF (Size: 510K)  | Related Articles | Citation Tracking

Identifiability of Finite Mixtures of Elliptical Distributions
HAJO HOLZMANN 1 , AXEL MUNKTILMANN GNEITING 2
  1 Institut für Mathematische Stochastik, University of Göttingen
  2 Department of Statistics, University of Washington
Correspondence to Hajo Holzmann, Institute for Mathematical Stochastics, Georg-August-Universität Göttingen, Maschmühlenweg 8–10, 37073 Göttingen, Germany.
E-mail: holzmann@math.uni-goettingen.de
Copyright 2006 Board of the Foundation of the Scandinavian Journal of Statistics.
KEYWORDS
characteristic function • elliptically symmetric • finite mixture • identifiability • Laplace transform • multivariate t distribution • normal scale mixture

ABSTRACT

Abstract. We present general results on the identifiability of finite mixtures of elliptical distributions under conditions on the characteristic generators or density generators. Examples include the multivariate t-distribution, symmetric stable laws, exponential power and Kotz distributions. In each case, the shape parameter is allowed to vary in the mixture, in addition to the location vector and the scatter matrix. Furthermore, we discuss the identifiability of finite mixtures of elliptical densities with generators that correspond to scale mixtures of normal distributions.


Received May 2005, in final form January 2006

DIGITAL OBJECT IDENTIFIER (DOI)
10.1111/j.1467-9469.2006.00505.x About DOI

Related Articles

  • Find other articles like this in Wiley InterScience
  • Find articles in Wiley InterScience written by any of the authors

Wiley InterScience is a member of CrossRef.

Cross Ref Member


Also of Interest

Statistics

Wiley-Blackwell is the largest publisher of society-based statistics journals and No. 1 in terms of quality and international scope.

Wiley-Blackwell publishes 19 statistics journals and is now the top publisher of Thomson Reuters ranked statistics journals.

Discover more about the statistics portfolio

Latest Impact Factor
Biometrics

Biometrics

Biometrics increased its
Impact Factor to: 1.970*

  • One of the top journals in Statistics and Probability
  • Highest Impact Factor Ever
  • Submit your paper today!

    * Thomson Reuters Journal Citation Report 2008

    Call for Papers
    Teaching Statistics

    Teaching Statistics

    Want to share your knowledge with those teaching pupils aged 9 – 19?

    Do you have a paper of interest to those teaching statistics, mathematics or economics?

    Submit Your Paper Today

    Hot Papers
    INSR

    International Statistical Review

    See the Papers attracting early citation:

    Econometric Causality

    Regression Revisited

    Hot Papers
    INSR

    Scandinavian Journal of Statistics

    See the Papers attracting early citation:

    Posterior Analysis for Normalized Random Measures with Independent Increments

    Bayesian Goodness of Fit Testing with Mixtures of Triangular Distributions

    Hot Papers
    INSR

    Scandinavian Journal of Statistics

    See the latest papers as soon as they publish online with EARLY VIEW

    Papers are fully copy-edited and peer-reviewed.

    Early View