Optimal Transportation and Applications
timetable:
Thu 5 Dec, 14:30 - 15:15, Aula Dini
<< Go back
Gradient flows in the geometry of Sinkhorn divergences
speaker: Hugo Lavenant (Università Bocconi, Milano)abstract: What happens to Wasserstein gradient flows if one uses entropic optimal transport in the JKO scheme instead of plain optimal transport? I will explain why it may be relevant to use Sinkhorn divergences, built on entropic optimal transport, as they allow the regularization parameter to remain fixed. This approach leads to studying the Riemannian geometry induced by Sinkhorn divergences, which retains some characteristics of the optimal transport geometry while being smoother. The gradient flows of potential energies in this geometry reveal intriguing features that I will discuss. This is joint work with Mathis Hardion, Jonas Luckhardt, Gilles Mordant, Bernhard Schmitzer, and Luca Tamanini.
timetable:
Thu 5 Dec, 14:30 - 15:15, Aula Dini
<< Go back