CMStatistics 2017: Start Registration
View Submission - CMStatistics
B0827
Title: Equivalent representations of max-stable processes via $\ell^p$ norms Authors:  Marco Oesting - University of Siegen (Germany) [presenting]
Abstract: Max-stable processes are popular models in spatial and spatio-temporal extremes. While they are typically written as pointwise maxima over an infinite number of stochastic processes, a family of novel representations based on $\ell^p$ norms is presented. This family includes both the construction of the Reich-Shaby model and the classical spectral representation by de Haan as special cases. As different representations may be advantageous for the purposes of statistical inference or simulation, formulae to switch between different equivalent representations are provided. Further, several properties of the represented processes are discussed.