Dealing with moment measures via entropy and optimal transport

Abstract

A recent paper by Cordero-Erausquin and Klartag provides a characterization of the measures $\mu$ on ${\mathbb{R}}^d$ which can be expressed as the moment measures of suitable convex functions $u$, i.e. are of the form $(\nabla u)\_\\#e^{- u}$ for $u:{\mathbb{R}}^d\to \mathbb{R}\cup \{+\infty \}$ and finds the corresponding $u$ by a variational method in the class of convex functions. Here we propose a purely optimal-transport-based method to retrieve the same result. The variational problem becomes the minimization of an entropy and a transport cost among densities $\rho$ and the optimizer $\rho$ turns out to be $e^{-u}$. This requires to develop some estimates and some semicontinuity results for the corresponding functionals which are natural in optimal transport. The notion of displacement convexity plays a crucial role in the characterization and uniqueness of the minimizers.