Dados do Trabalho

Título

AN EVALUATION OF DIFFERENT GUMBEL PARAMETER ESTIMATION APPROACHES

Resumo

The maximum likelihood method and the method of moments are commonly used for Gumbel parameter
estimation.The maximum likelihood estimator cannot be expressed in closed-form. Estimates are
typically obtained by numerically maximizing the log likelihood. An alternative estimation method
which is less well-known is the probability weighted moments method. The main goal of this paper
is to compare the performances of such estimation methods. To do this, we used Monte Carlo
simulation on Gumbel distribution and evaluate comparative measures based on average estimates,
average biases and average standard errors using parametric bootstrap replications.

Palavras-chave

Estimation methods, Probability weighted moments method, Maximum likelihood method, Method of moments, Monte Carlo simulation