Title: Multivariate Bernoulli copulas: properties and limitations
Authors: Elisa Perrone - Eindhoven University of Technology (Netherlands) [presenting]
Roberto Fontana - Politecnico di Torino (Italy)
Abstract: Possible extensions to higher dimensions of the notion of Bernoulli copula are analyzed. We show that there are various ways to define Bernoulli copulas in dimension three (and higher), and that the choice is generally arbitrary. We investigate the pros and cons of these extensions and inquire into the uniqueness of the associated statistical models in relation to additional stochastic constraints, such as exchangeability.