English

Estimating causal distances with non-causal ones

Probability 2025-10-24 v2

Abstract

The adapted Wasserstein (AWAW) distance refines the classical Wasserstein (WW) distance by incorporating the temporal structure of stochastic processes. This makes the AWAW-distance well-suited as a robust distance for many dynamic stochastic optimization problems where the classical WW-distance fails. However, estimating the AWAW-distance is a notably challenging task, compared to the classical WW-distance. In the present work, we build a sharp estimate for the AWAW-distance in terms of the WW-distance, for smooth measures. This reduces estimating the AWAW-distance to estimating the WW-distance, where many well-established classical results can be leveraged. As an application, we prove a fast convergence rate of the kernel-based empirical estimator under the AWAW-distance, which approaches the Monte-Carlo rate (n1/2n^{-1/2}) in the regime of highly regular densities. These results are accomplished by deriving a sharp bi-Lipschitz estimate of the adapted total variation distance by the classical total variation distance.

Keywords

Cite

@article{arxiv.2506.22421,
  title  = {Estimating causal distances with non-causal ones},
  author = {Beatrice Acciaio and Songyan Hou and Gudmund Pammer},
  journal= {arXiv preprint arXiv:2506.22421},
  year   = {2025}
}
R2 v1 2026-07-01T03:36:55.378Z