CMStatistics 2018: Start Registration
View Submission - CMStatistics
Title: An improved prior choice for the parameters in the generalized Pareto distribution Authors:  Eva Lopez Sanjuan - Universidad de Extremadura (Spain) [presenting]
Mario Martinez Pizarro - University of Extremadura (Spain)
M Isabel Parra Arevalo - Universidad de Extremadura (Spain)
Jacinto Martin Jimenez - Universidad de Extremadura (Spain)
Abstract: In the parameter estimation of limit extreme value distributions, standard methods only use some of the available data. For the generalized Pareto distribution, only the observations above a certain threshold are considered, therefore a big amount of information is wasted. The aim is making the most of the information provided by the observations, in order to improve the accuracy in Bayesian parameter estimation. The strategy consists in taking advantage of the existing relationship between the parameters of baseline and generalized Pareto distributions to obtain informative prior distributions. Different simulations have been carried out in order to compare the effectiveness of the proposed method to the standard ones. Specifically, simulations for different baseline distributions were studied: normal, exponential and Cauchy distributions, because of the different behavior of the tails. The proposed method can be extended to other baseline distributions.