28
9

Central Limit Theorems for Semidiscrete Wasserstein Distances

Abstract

We prove a Central Limit Theorem for the empirical optimal transport cost, nmn+m{Tc(Pn,Qm)Tc(P,Q)}\sqrt{\frac{nm}{n+m}}\{\mathcal{T}_c(P_n,Q_m)-\mathcal{T}_c(P,Q)\}, in the semi discrete case, i.e when the distribution PP is supported in NN points, but without assumptions on QQ. We show that the asymptotic distribution is the supremun of a centered Gaussian process, which is Gaussian under some additional conditions on the probability QQ and on the cost. Such results imply the central limit theorem for the pp-Wassertein distance, for p1p\geq 1. This means that, for fixed NN, the curse of dimensionality is avoided. To better understand the influence of such NN, we provide bounds of EW1(P,Qm)W1(P,Q)E|\mathcal{W}_1(P,Q_m)-\mathcal{W}_1(P,Q)| depending on mm and NN. Finally, the semidiscrete framework provides a control on the second derivative of the dual formulation, which yields the first central limit theorem for the optimal transport potentials. The results are supported by simulations that help to visualize the given limits and bounds. We analyse also the cases where classical bootstrap works.

View on arXiv
Comments on this paper