Estimation of parameters in a finite mixture of multivariate gamma distributions using gaussian approximation

Finite mixture of multivariate gamma distributions is extensively used in the domains of stochastic modelling, reliability, hydrology and life testing. In this paper, we consider a multivariate gamma mixture model (MGMM) with independent marginals. A novel approach is proposed for estimating the parameters of this model. The approach makes use of Wilson-Hilferty approximation, MCLUST algorithm and the principle of maximum likelihood. Numerical illustrations based on simulated as well as real datasets have been implemented to assess the performance of the proposed approach. The results indicate that the proposed methodology provides reliable estimates for the model parameters.

This work is licensed under a Creative Commons Attribution 4.0 International License.