Comparación del método del principio de la máxima entropía en la estimación de parámetros de la distribución de valores extremos tipo I

Abstract

The method of the Principle of Maximum Entropy (POME) applied to the estimation of parameters of the extreme value type I distribution, (EVI) is analyzed. The POME method has been compared with others of widespread use, like the methods of moments (MOM), maximum likelihood (ML) and probability weighted moments (PWM), with both real flood data and through distributional sampling experiments. The POME method was another good option for estimating the parameters of the EVI distribution, but not as good as those provided by the methods of PWM, ML and MOM. It was also detected that the POME method has a better performance when the sample size is bigger than 50 values of maximum annual floods.