It has been conjectured by Khatri and Mardia that with probability one MLEs for the parameters of the von Mises-Fisher matrix distribution exist and are unique. We prove that, except for small sample sizes, this conjecture is true, both in the case where the parameter matrix has known rank and in the unrestricted case. The corresponding result for the matrix Bingham distribution is proven also.
Publié le : 1979-05-14
Classification:
Maximum likelihood estimator,
von Mises-Fisher matrix distribution,
Bingham matrix distribution,
exponential family,
62F10,
62F05
@article{1176344681,
author = {Jupp, P. E. and Mardia, K. V.},
title = {Maximum Likelihood Estimators for the Matrix Von Mises-Fisher and Bingham Distributions},
journal = {Ann. Statist.},
volume = {7},
number = {1},
year = {1979},
pages = { 599-606},
language = {en},
url = {http://dml.mathdoc.fr/item/1176344681}
}
Jupp, P. E.; Mardia, K. V. Maximum Likelihood Estimators for the Matrix Von Mises-Fisher and Bingham Distributions. Ann. Statist., Tome 7 (1979) no. 1, pp. 599-606. http://gdmltest.u-ga.fr/item/1176344681/