On a bivariate generalized gamma distribution

Abstract

A bivariate generalized gamma distribution (with marginal distributions of a gamma generalized type) is obtained from a bivariate normal distribution by means of a transformation. The analytic form of the joint density function presents symmetry between the marginal variables and a single non linear correlation parameter. Marginal and conditional densities distributions are obtained at first and, next, the expression of the moments. The estimation of the parameters is briefly dealt too.