On a Bivariate XGamma Distribution Derived from Copula

Abstract

In this paper, a new bivariate XGamma (BXG) distribution is presented using Farlie-Gumbel-Morgenstern (FGM) copula. We derive the expressions for conditional distribution, regression function and product moments for the BXG distribution. Concept of reliability and various measures of local dependence are also studied for the proposed model. Furthermore, estimation of the parameters of the BXG distribution is obtained through maximum likelihood estimation and inference function of margin estimation procedures. Finally, an application of the same is also demonstrated to a real data set.

Keywords

• bivariate xgamma distribution
• copulas
• fgm copula
• maximum likelihood estimate
• inference function of margin