Comparison of Linear Regression and Maximum Likelihood Methods to Estimate a New Pareto-type Distribution

This paper compares methods of estimation for the parameters of a New Pareto-type distribution. Least-squares (LS) regression and maximum likelihood estimates (MLE) are presented. For the LS regression method, LS Y on X and LS X on Y are used for estimating the two parameters for the new Pareto-type distribution. The properties of LS regression and maximum likelihood estimators are compared via Monte Carlo simulations. We also compare the two LS regression methods LS Y on X and LS X on Y in terms of model statistics.

Keywords:

Maximum likelihood estimator , least square estimator , mean square error.