Foto sezione
Logo Bocconi

Events of Department of Decision Sciences

May 14, 2020 at 16:30


Mappings valued in the Wasserstein space

Hugo Lavenant (University of British Columbia)

Abstract. The so-called Wasserstein space is the space of probability distributions over a fixed domain endowed with the Wasserstein distance coming from optimal transport. It can be seen as an infinite dimensional Riemannian manifold. In this talk, I will give a brief introduction to Wasserstein spaces and (some of) their many applications. Then, I will focus on a definition of Dirichlet energy for mappings valued in the Wasserstein space, whose minimizers are harmonic mappings. I will present one recent application of such a concept by using it to build a convex relaxation of a problem of nonlinear elasticity.


by invitation: for information or to receive the invitation link contact

In this section