Authors: Célestin C. Kokonendji and Khoirin Nisa
Abstract: This paper presents three estimations of generalized
variance (i.e. determinant of covariance matrix) of normal-
Poisson models: maximum likelihood (ML) estimator, uniformly
minimum variance unbiased (UMVU) estimator, and Bayesian
estimator. First, the definition and some properties of normal
Poisson models are established. Then ML, UMVU and Bayesian
estimators for generalized variance are derived. Finally, a
simulation study is carried out to assess the performance of the
estimators based on their mean square error (MSE).
Keywords: Covariance matrix, determinant, normal stable Tweedie, maximum likelihood, UMVU, Bayesian estimator.