Some parametric models on the simplex
A class of new parametric models on the unit simplex in Rm is introduced, the distributions in question being obtained as conditional distributions of m independent generalized inverse Gaussian random variables given their sum. The Dirichlet model occurs as a special case. Two other special cases, corresponding respectively to the inverse Gaussian model and the reciprocal inverse Gaussian model, are studied in some detail. In particular, several exact chi-squared decompositions are found.
Year of publication: |
1991
|
---|---|
Authors: | Barndorff-Nielsen, O. E. ; Jørgensen, B. |
Published in: |
Journal of Multivariate Analysis. - Elsevier, ISSN 0047-259X. - Vol. 39.1991, 1, p. 106-116
|
Publisher: |
Elsevier |
Keywords: | chi-squared decompositions compositional data Dirichlet distribution exponential models generalized inverse Gaussian distribution inverse Gaussian distribution reciprocal inverse Gaussian distribution |
Saved in:
Saved in favorites
Similar items by person
-
WORKSHOP - Fitting Tweedie's Compound Poisson Model to Insurance Claims Data: Dispersion Modelling
Smyth, G.K., (2002)
-
On conditionality resolution and the likelihood ratio for curved exponential models
Barndorff-Nielsen, O. E., (1984)
-
Tail Exactness of Multivariate Saddlepoint Approximations
Barndorff-Nielsen, O. E., (1999)
- More ...