Dense Classes of Multivariate Max Stable Distributions


  • Profª Cécile Mercadier – Université de Lyon; CNRS – FRANCE
  • FCUL (DEIO) – Campo Grande – Bloco C8 Piso 2 – Sala 8.2.23 – 14:30h
  • Sexta-feira, 25 de Novembro de 2011

Our goal is to describe flexible max-stable models in high dimensions. The first class is the well-known max stable distributions with discrete angular measures. It is tractablein all dimensions, with simple exact methods to simulate and compute the distributionfunction. The second class is max stable models with piecewise polynomial density for the angular measure. Currently, these are only accessible in two dimensions. A third class of models is the family of generalized asymmetric logistic models. We detail the connection between this class of models and the closely related class of generalized stable mixtures.

It is shown that all three classes are dense in the space of max stable models.

This is a collaboration with Anne-Laure Fougères (Université de Lyon) and John P. Nolan (American University).