Some variations of EM algorithms for Marshall-Olkin bivariate Pareto distribution with location and scale

TL;DR

This paper explores alternative EM algorithms for estimating the parameters of the Marshall-Olkin bivariate Pareto distribution, validating their performance through simulations and real data analysis.

Contribution

It introduces new variants of the EM algorithm tailored for the seven-parameter Marshall-Olkin bivariate Pareto distribution, expanding estimation methods.

Findings

Alternative EM algorithms perform well in simulations

Proposed methods provide accurate parameter estimates

Real data analysis demonstrates practical applicability

Abstract

Recently Asimit et. al used an EM algorithm to estimate Marshall-Olkin bivariate Pareto distribution. The distribution has seven parameters. We describe few alternative approaches of EM algorithm. A numerical simulation is performed to verify the performance of different proposed algorithms. A real-life data analysis is also shown for illustrative purposes.