报告题目 (Title)：Convergence rate of Smoothed empirical Wasserstein distance
（平滑经验Wasserstein距离的收敛速度）
报告人 (Speaker)：黎怀谦 教授（天津大学）
报告时间 (Time)：2025年11月14日（周五）16：00-19:00
报告地点 (Place)： 腾讯会议：880-966-912
邀请人(Inviter)：阳芬芬
主办部门：理学院数学系
报告摘要：This work provides upper bounds for the expected Gaussian-smoothed image -Wasserstein distance. We analyze the convergence rate of the empirical measure image , constructed from image samples, towards the true probability measure image on image . The bounds hold universally for any image —encompassing the total variation distance when image —under mild moment assumptions on image . Furthermore, we extend our analysis beyond the independent and identically distributed case, deriving convergence results for dependent data generated by image -mixing sequences and Markov chains.
