Resumen

FLOWERS-CANO, Roberto S.; FLOWERS, Robert Jeffrey  y  RIVERA-TREJO, Fabián. Evaluation of Probabilistic Model Selection Criteria: Validation with Series of Simulated Maxima. Tecnol. cienc. agua [online]. 2014, vol.5, n.5, pp.189-197. ISSN 2007-2422.

In this paper, a Monte Carlo study is performed to determine the validity of using the root mean square error as a selection criterion for the frequency analysis. This statistic is compared to those from the Kolmogorov- Smirnov, the Cramer-Von Mises and the Anderson-Darling tests. The distributions chosen to compare these statistics are gamma, Weibull, Gumbel, log-normal and log-logistic. The results obtained by this study indicate that at least 50 random observations are needed to obtain good results using either the Anderson-Darling or root mean square error statistics. The Kolmogorov-Smirnov and Cramer-Von Mises tests are not recommended since they require sample sizes larger than those usually found in hydrology.

Palabras llave : Root mean square error; Kolmogorov- Smirnov; Cramer-Von Mises; Anderson-Darling.

        · resumen en Español     · texto en Español     · Español ( pdf )