Exponential convergence in L^p-Wasserstein distance for diffusion processes without uniformly dissipative drift

DOI10.1002/MANA.201500351zbMATH Open1350.60079arXiv1407.1986OpenAlexW2963730516MaRDI QIDQ2830673FDOQ2830673

Authors: Jian Wang, Dejun Luo 

Publication date: 28 October 2016

Published in: Mathematische Nachrichten (Search for Journal in Brave)

Abstract: By adopting the coupling by reflection and choosing an auxiliary function which is convex near infinity, we establish the exponential convergence of diffusion semigroups $(P_t{)}_{tge0}$ with respect to the standard $L^p$-Wasserstein distance for all $pin[1,infty)$. In particular, we show that for the It^o stochastic differential equation ${\displaystyle }$d X_t=d B_t+b(X_t),d t,${\displaystyle }$ if the drift term $b$ satisfies that for any $x,yin{R}^d$, ${\displaystyle }$langle b(x)-b(y),x-y anglele �egin{cases} K_1|x-y|^2,& |x-y|le L; -K_2|x-y|^2,& |x-y|> L end{cases}${\displaystyle }$ holds with some positive constants $K_1$, $K_2$ and $L>0$, then there is a constant $lambda:=lambda(K_1,K_2,L)>0$ such that for all $pin[1,infty)$, $t>0$ and $x,yin{R}^d$, ${\displaystyle }$W_p(delta_x P_t,delta_y P_t)leq Ce^{-lambda t/p} �egin{cases} |x-y|^{1/p}, & mbox{if } |x-y|le 1; |x-y|, & mbox{if } |x-y|> 1. end{cases}${\displaystyle }$ where $C:=C(K_1,K_2,L,p)$ is a positive constant. This improves the main result in cite{Eberle} where the exponential convergence is only proved for the $L^1$-Wasserstein distance.

Full work available at URL: https://arxiv.org/abs/1407.1986

Recommendations

• Exponential convergence in the Wasserstein metric \(W_1\) for one dimensional diffusions
• \(L^{p}\)-Wasserstein distance for stochastic differential equations driven by Lévy processes
• Reflection coupling and Wasserstein contractivity without convexity
• Refined basic couplings and Wasserstein-type distances for SDEs with Lévy noises
• Exponential contraction in Wasserstein distances for diffusion semigroups with negative curvature

zbMATH Keywords

couplingreflectionstochastic differential equationsdiffusion processesexponential convergence\(L^p\)-Wasserstein distance

Mathematics Subject Classification ID

Diffusion processes (60J60) Stochastic ordinary differential equations (aspects of stochastic analysis) (60H10)

Cites Work

• Markov chains and stochastic stability
• Gradient estimates for diffusion semigroups with singular coefficients
• Well-posedness of the transport equation by stochastic perturbation
• Rate of convergence for ergodic continuous Markov processes: Lyapunov versus Poincaré
• Stability of Markovian processes III: Foster–Lyapunov criteria for continuous-time processes
• Functional inequalities for empty essential spectrum
• Title not available (Why is that?)
• Weak Poincaré inequalities and \(L^2\)-convergence rates of Markov semigroups
• Coupling of multidimensional diffusions by reflection
• Coupling methods for multidimensional diffusion processes
• Convergence to equilibrium in Wasserstein distance for Fokker-Planck equations
• Transport inequalities, gradient estimates, entropy and Ricci curvature
• Estimation of spectral gap for elliptic operators
• A note on Talagrand's transportation inequality and logarithmic Sobolev inequality
• Reflection coupling and Wasserstein contractivity without convexity
• Spectral gaps in Wasserstein distances and the 2D stochastic Navier-Stokes equations
• Basic estimates of stability rate for one-dimensional diffusions
• Subgeometric ergodicity of strong Markov processes
• Asymptotic coupling and a general form of Harris' theorem with applications to stochastic delay equations
• Uniform convergence to equilibrium for granular media
• Subgeometric rates of convergence of Markov processes in the Wasserstein metric
• Existence of the spectral gap for elliptic operators

Cited In (28)

• Convergence in Monge-Wasserstein distance of mean field systems with locally Lipschitz coefficients
• Exponential convergence for functional SDEs with Hölder continuous drift
• Exponential ergodicity for stochastic functional differential equations with Markovian switching
• Asymptotics of smoothed Wasserstein distances
• Exponential contraction in Wasserstein distances for diffusion semigroups with negative curvature
• Ergodicity of the underdamped mean-field Langevin dynamics
• Evolution of the Wasserstein distance between the marginals of two Markov processes
• Subexponential upper and lower bounds in Wasserstein distance for Markov processes
• Precise limit in Wasserstein distance for conditional empirical measures of Dirichlet diffusion processes
• Convergence in Wasserstein distance for self-stabilizing diffusion evolving in a double-well landscape
• On exponential decay of a distance between solutions of an SDE with non-regular drift
• Uniqueness of stationary distribution and exponential convergence for distribution dependent SDEs
• Exponential ergodicity for general continuous-state nonlinear branching processes
• Invariant probability measures for path-dependent random diffusions
• Refined basic couplings and Wasserstein-type distances for SDEs with Lévy noises
• \(L^{p}\)-Wasserstein distance for stochastic differential equations driven by Lévy processes
• Exponential convergence in the Wasserstein metric \(W_1\) for one dimensional diffusions
• Wasserstein contraction and Poincaré inequalities for elliptic diffusions with high diffusivity
• Strict Kantorovich contractions for Markov chains and Euler schemes with general noise
• Convergence in variation of solutions of nonlinear Fokker-Planck-Kolmogorov equations to stationary measures
• Gradient estimates and ergodicity for SDEs driven by multiplicative Lévy noises via coupling
• Nonasymptotic bounds for sampling algorithms without log-concavity
• On sub-geometric ergodicity of diffusion processes
• \(L^2\)-Wasserstein contraction for Euler schemes of elliptic diffusions and interacting particle systems
• Sub-exponential convergence to equilibrium for Gaussian driven stochastic differential equations with semi-contractive drift
• Unadjusted Langevin algorithm with multiplicative noise: total variation and Wasserstein bounds
• Maximum entropy methods for texture synthesis: theory and practice
• Long-time behaviors of mean-field interacting particle systems related to McKean-Vlasov equations

This page was built for publication: Exponential convergence in \(L^p\)-Wasserstein distance for diffusion processes without uniformly dissipative drift

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2830673)