English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

A major issue of extreme value analysis is the determination of the shapeparameter $\xi$ common to Generalized Extreme Value (GEV) and GeneralizedPareto (GP) distributions, which drives the tail behavior, and is of majorimpact on the estimation of return levels and periods. Many practitioners makethe choice of a Bayesian framework to conduct this assessment for accounting ofparametric uncertainties, which are typically high in such analysescharacterized by a low number of observations. Nonetheless, such approaches canprovide large credibility domains for $\xi$, including negative and positivevalues, which does not allow to conclude on the nature of the tail. Consideringthe block maxima framework, a generic approach of the determination of thevalue and sign of $\xi$ arises from model selection between the Fr\'echet,Gumbel and Weibull possible domains of attraction conditionally toobservations. Opposite to the common choice of the GEV as an appropriate modelfor {\it sampling} extreme values, this model selection must be conducted withgreat care. The elicitation of proper, informative and easy-to use priors isconducted based on the following principle: for all parameter dimensions theyact as posteriors of noninformative priors and virtual samples. Statistics ofthese virtual samples can be assessed from prior predictive information, and acompatibility rule can be carried out to complete the calibration, even thoughthey are only semi-conjugated. Besides, the model selection is conducted usinga mixture encompassing framework, which allows to tackle the computation ofBayes factors. Motivating by a real case-study involving the elicitation ofexpert knowledge on meteorological magnitudes, the overall methodology isillustrated by toy examples too.

Bayesian prior elicitation and selection for extreme values