Parameter Estimation for the Two-Parameter Maxwell Distribution Under Complete and Censored Samples

Abstract

The Maxwell distribution is one of the basic distributions in Physics besides being popular in Statistics for modeling lifetime data. This paper considers the parameter estimation of the Maxwell distribution via modified maximum likelihood (MML) methodology for both complete and censored samples. The MML estimators for the location and scale parameters of the Maxwell distribution have explicit forms and they are robust against the plausible deviations from the assumed model. A Monte Carlo simulation study is conducted to compare the performances of the MML estimators with the corresponding maximum likelihood (ML), least squares (LS) and method of moments (MoM) estimators.

Keywords

• efficiency
• Maxwell distribution
• modified likelihood
• Monte Carlo simulation
• Newton–Raphson
• Type-II censoring