A UNIFIED APPROACH TO MAXIMUM LIKELIHOOD ESTIMATION

In this paper a unified approach to maximum likelihood estimation (MLE) of parameters of statistical distributions by the sequential number theoretic algorithm for optimization (SNTO) proposed by Fang and Wang (1989) is adopted. For illustration of the use of this approach more attention is paid to MLE's of parameters of the Weibull and the beta distributions. A number of examples show that the present method is universally useful and effective.