Aller au contenu Aller au menu Aller à la recherche

Événements et séminaires - LSTA
Laboratoire de Statistique Théorique et Appliquée


22/05/2014 - Guilia Marcon (Bocconi)

Groupe de travail Théorie des valeurs extrêmes

Nonparametric Estimation of the Pickands Dependence Function using Bernstein Polynomials

Many applications require the estimation of the dependence of high dimensional extreme events. A way to do this is by estimating the so called Pickands dependence function. While the advantage of a nonparametric approach is the flexibility in capturing the dependence, in high dimensions, i.e. greater than two, the estimation is a challenging task. In this talk we propose a nonparametric approach for estimating the Pickands dependence function, consisting of the Bernstein polynomials representation of a new multivariate nonparametric estimator. We discuss the properties of the proposed estimation method and illustrate its performance with a simulation study. We show the applicability of this framework in high-dimensional problems, analyzing the seventh- dimensional dependence structure of heavy rainfall in France. It is a joint work with S. A. Padoan, P. Naveau and P. Muliere.