Identifiability of Finite Mixtures of Elliptical Distributions
- 17 March 2006
- journal article
- research article
- Published by Wiley in Scandinavian Journal of Statistics
- Vol. 33 (4) , 753-763
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.Keywords
This publication has 20 references indexed in Scilit:
- A Report on the Future of StatisticsStatistical Science, 2004
- Model-Based Clustering, Discriminant Analysis, and Density EstimationJournal of the American Statistical Association, 2002
- Computer-Assisted Analysis of Mixtures and ApplicationsTechnometrics, 2000
- Consistent Estimation of a Mixing DistributionThe Annals of Statistics, 1992
- Identifiability of Finite Mixtures for Directional DataThe Annals of Statistics, 1983
- On the identifiability of finite mixtures of distributions (Corresp.)IEEE Transactions on Information Theory, 1981
- Generalized Hyperbolic Secant DistributionsJournal of the American Statistical Association, 1968
- On the Identifiability of Finite MixturesThe Annals of Mathematical Statistics, 1968
- Identifiability of Finite MixturesThe Annals of Mathematical Statistics, 1963
- Identifiability of MixturesThe Annals of Mathematical Statistics, 1961