Exponential Contraction in Wasserstein Distances for Diffusion Semigroups with Negative Curvature

被引:0
|
作者
Feng-Yu Wang
机构
[1] Beijing Normal University,Laboratory of Mathematical and Complex Systems
[2] Swansea University,Department of Mathematics
来源
Potential Analysis | 2020年 / 53卷
关键词
Wasserstein distance; Diffusion semigroup; Riemannian manifold; Curvature condition; SDEs with multiplicative noise; 60J75; 47G20; 60G52;
D O I
暂无
中图分类号
学科分类号
摘要
Let Pt be the (Neumann) diffusion semigroup Pt generated by a weighted Laplacian on a complete connected Riemannian manifold M without boundary or with a convex boundary. It is well known that the Bakry-Emery curvature is bounded below by a positive constant ≪> 0 if and only if Wp(μ1Pt,μ2Pt)≤e−≪tWp(μ1,μ2),t≥0,p≥1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$W_{p}(\mu_{1}P_{t}, \mu_{2}P_{t})\leq e^{-\ll t} W_{p} (\mu_{1},\mu_{2}),\ \ t\geq 0, p\geq 1 $$\end{document} holds for all probability measures μ1 and μ2 on M, where Wp is the Lp Wasserstein distance induced by the Riemannian distance. In this paper, we prove the exponential contraction Wp(μ1Pt,μ2Pt)≤ce−≪tWp(μ1,μ2),p≥1,t≥0\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$W_{p}(\mu_{1}P_{t}, \mu_{2}P_{t})\leq ce^{-\ll t} W_{p} (\mu_{1},\mu_{2}),\ \ p \geq 1, t\geq 0$$\end{document} for some constants c,≪> 0 for a class of diffusion semigroups with negative curvature where the constant c is essentially larger than 1. Similar results are derived for SDEs with multiplicative noise by using explicit conditions on the coefficients, which are new even for SDEs with additive noise.
引用
收藏
页码:1123 / 1144
页数:21
相关论文
共 25 条
  • [1] Exponential Contraction in Wasserstein Distances for Diffusion Semigroups with Negative Curvature
    Wang, Feng-Yu
    POTENTIAL ANALYSIS, 2020, 53 (03) : 1123 - 1144
  • [3] Dimensional Contraction in Wasserstein Distance for Diffusion Semigroups on a Riemannian Manifold
    Ivan Gentil
    Potential Analysis, 2015, 42 : 861 - 873
  • [4] Graph Diffusion Wasserstein Distances
    Barbe, Amelie
    Sebban, Marc
    Goncalves, Paulo
    Borgnat, Pierre
    Gribonval, Remi
    MACHINE LEARNING AND KNOWLEDGE DISCOVERY IN DATABASES, ECML PKDD 2020, PT II, 2021, 12458 : 577 - 592
  • [5] On exponential stability of contraction semigroups
    Carlos S. Kubrusly
    Nhan Levan
    Semigroup Forum, 2011, 83 : 513 - 521
  • [6] On exponential stability of contraction semigroups
    Kubrusly, Carlos S.
    Nhan Levan
    SEMIGROUP FORUM, 2011, 83 (03) : 513 - 521
  • [7] EXPONENTIAL CONTRACTION IN WASSERSTEIN DISTANCE ON STATIC AND EVOLVING MANIFOLDS
    Cheng, Li-Juan
    Thalmaier, Anton
    Zhang, Shao-Qin
    REVUE ROUMAINE DE MATHEMATIQUES PURES ET APPLIQUEES, 2021, 66 (01): : 107 - 129
  • [8] STOCHASTIC EQUATION AND EXPONENTIAL ERGODICITY IN WASSERSTEIN DISTANCES FOR AFFINE PROCESSES
    Friesen, Martin
    Jin, Peng
    Rudiger, Barbara
    ANNALS OF APPLIED PROBABILITY, 2020, 30 (05): : 2165 - 2195
  • [10] Nonlinear Diffusion: Geodesic Convexity is Equivalent to Wasserstein Contraction
    Bolley, Francois
    Carrillo, Jose A.
    COMMUNICATIONS IN PARTIAL DIFFERENTIAL EQUATIONS, 2014, 39 (10) : 1860 - 1869