ON GRADIENT STRUCTURES FOR MARKOV CHAINS AND THE PASSAGE TO WASSERSTEIN GRADIENT FLOWS

被引:19
|
作者
Disser, Karoline [1 ]
Liero, Matthias [1 ]
机构
[1] Weierstrass Inst, D-10117 Berlin, Germany
基金
欧洲研究理事会;
关键词
Wasserstein gradient flow; discrete gradient flow structures; entropy/entropy-dissipation formulation; evolutionary Gamma-convergence; Markov chains; CONVERGENCE; DIFFUSION; EQUATIONS; EVOLUTION;
D O I
10.3934/nhm.2015.10.233
中图分类号
O1 [数学];
学科分类号
0701 ; 070101 ;
摘要
We study the approximation of Wasserstein gradient structures by their finite-dimensional analog. We show that simple finite-volume discretizations of the linear Fokker-Planck equation exhibit the recently established entropic gradient-flow structure for reversible Markov chains. Then we reprove the convergence of the discrete scheme in the limit of vanishing mesh size using only the involved gradient-flow structures. In particular, we make no use of the linearity of the equations nor of the fact that the Fokker-Planck equation is of second order.
引用
收藏
页码:233 / 253
页数:21
相关论文
共 50 条
  • [1] Gradient flows of the entropy for finite Markov chains
    Maas, Jan
    JOURNAL OF FUNCTIONAL ANALYSIS, 2011, 261 (08) : 2250 - 2292
  • [2] Characterisation of gradient flows on finite state Markov chains
    Dietert, Helge
    ELECTRONIC COMMUNICATIONS IN PROBABILITY, 2015, 20 : 1 - 8
  • [3] Approximate Inference with Wasserstein Gradient Flows
    Frogner, Charlie
    Poggio, Tomaso
    INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND STATISTICS, VOL 108, 2020, 108 : 2581 - 2589
  • [4] Policy Optimization as Wasserstein Gradient Flows
    Zhang, Ruiyi
    Chen, Changyou
    Li, Chunyuan
    Carin, Lawrence
    INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 80, 2018, 80
  • [5] Entropic Approximation of Wasserstein Gradient Flows
    Peyre, Gabriel
    SIAM JOURNAL ON IMAGING SCIENCES, 2015, 8 (04): : 2323 - 2351
  • [6] Primal Dual Methods for Wasserstein Gradient Flows
    Carrillo, Jose A.
    Craig, Katy
    Wang, Li
    Wei, Chaozhen
    FOUNDATIONS OF COMPUTATIONAL MATHEMATICS, 2022, 22 (02) : 389 - 443
  • [7] {Euclidean, metric, and Wasserstein} gradient flows: an overview
    Santambrogio, Filippo
    BULLETIN OF MATHEMATICAL SCIENCES, 2017, 7 (01) : 87 - 154
  • [8] Primal Dual Methods for Wasserstein Gradient Flows
    José A. Carrillo
    Katy Craig
    Li Wang
    Chaozhen Wei
    Foundations of Computational Mathematics, 2022, 22 : 389 - 443
  • [9] Variational inference via Wasserstein gradient flows
    Lambert, Marc
    Chewi, Sinho
    Bach, Francis
    Bonnabel, Silvere
    Rigollet, Philippe
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 35 (NEURIPS 2022), 2022,
  • [10] GRADIENT FLOWS FOR PROBABILISTIC FRAME POTENTIALS IN THE WASSERSTEIN SPACE
    Wickman, Clare
    Okoudjou, Kasso A.
    SIAM JOURNAL ON MATHEMATICAL ANALYSIS, 2023, 55 (03) : 2324 - 2346