AI Paper Atlas

Research map / Learning theory and algorithms / Optimal transport

Wasserstein F-tests for Frechet regression on Bures-Wasserstein manifolds

Haoshu Xu, Hongzhe Li

JMLR 2025 · Learning theory and algorithms › Optimal transport · approach: Wasserstein barycenters and GW

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use the proposed Wasserstein F-test instead of the distance covariance testdcov independence test · Simulated covariance-matrix regression, n=200, p in {1,3,5}test powerImproved
estimate covariance matrices from 50 or 100 samples instead of observing them directly (ablation)Wasserstein F-test · Simulation Example 1, n=200, p=5, d=5test power: n-tilde = infinity -> n-tilde in {50, 100}Hurt
estimate covariance matrices from 50 or 100 samples instead of observing them directly (ablation)Frechet regression estimator · Simulation Example 1, diagonal entries, n=200estimator bias: n-tilde = infinity -> n-tilde in {50, 100}Hurt
increase covariate dimension p and matrix dimension d (ablation)Wasserstein F-test statistic · Simulation Examples 1 and 2, n=200test statistic magnitude: p in {1,3,5}, d in {5,10}Mixed
initialize gradient descent at identity matrix instead of random or mean initialization (ablation)Riemannian gradient descent · Appendix simulationsconvergenceNo clear effect
use step size 1 instead of smaller step sizes (ablation)Riemannian gradient descent · Appendix simulationsconvergenceImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy