B1057
Title: On a generalized closed-form maximum likelihood estimator for some survival distributions
Authors: Francisco Louzada - University of Sao Paulo (Brazil) [presenting]
Abstract: Nowadays, estimates and predictions need to be computed in real time, especially in applications utilizing embedded technology. A generalization of the maximum likelihood estimator is introduced, which enables the obtention of estimators in closed-form expressions while outperforming the existing estimation procedures. The numerical findings demonstrate that the approach yields nearly unbiased estimates even for small sample sizes. The proposed generalized version of the maximum likelihood estimator is illustrated on the Nakagami-type distributions, which play an important role in communication engineering problems, particularly in the model-fading of radio signals. Furthermore, the applicability of the methodology is showcased to other probability distributions, such as gamma and generalized gamma distributions.
