Processus max-stables

Clément Dombry (Université de Franche-Comté)

A guided tour in the theory of max-stable processes

The theory and applications of max-stable processes have been developing rapidly over the last decades. Max-stable processes are important models in spatial extreme value theory since they arise as the only possible non-degenerate limits of maxima of iid processes under affine normalization. They are used to assess environmental  risks in the framework of the annual maxima method. The purpose of these lectures is to provide an introduction to the theory of max-stable processes with a particular focus on a series of recent results obtained thanks to a careful analysis of de Haan's Poisson point process representation of max-stable processes.

Tentative program: 

  1. Fundamental grounds of the theory
  2. Models for max-stable processes
  3. Extremal functions, subextremal functions and hitting scenario
  4. Exact simulation and conditional simulation
  5. Pairwise likelihood and full likelihood based inference
  6. Probability of concurrent extremes

Date Cours Lieu
11 May 2016 10:00 Processus max-stables Université Pierre et Marie Curie, Salle 24-34-201
12 May 2016 10:00 Processus max-stables Université Pierre et Marie Curie, Salle 24-34-201
18 May 2016 10:00 Processus max-stables Université Pierre et Marie Curie, salle 15-16-201
19 May 2016 10:00 Processus max-stables Université Pierre et Marie Curie, salle 15-16-201