Multiscale integrators for stochastic differential equations and irreversible Langevin samplers

We study multiscale integrator numerical schemes for a class of stiff stochastic differential equations (SDEs). We consider multiscale SDEs that behave as diffusions on graphs as the stiffness parameter goes to its limit. Classical numerical discretization schemes, such as the Euler-Maruyama scheme, become unstable as the stiffness parameter converges to its limit and appropriate multiscale integrators can correct for this. We rigorously establish the convergence of the numerical method to the related diffusion on graph, identifying the appropriate choice of discretization parameters. Theoretical results are supplemented by numerical studies on the problem of the recently developing area of introducing irreversibility in Langevin samplers in order to accelerate convergence to equilibrium.

[1]  G. Pavliotis,et al.  Variance Reduction Using Nonreversible Langevin Samplers , 2015, Journal of statistical physics.

[2]  K. Spiliopoulos,et al.  Irreversible Langevin samplers and variance reduction: a large deviations approach , 2014, 1404.0105.

[3]  Assyr Abdulle,et al.  High Weak Order Methods for Stochastic Differential Equations Based on Modified Equations , 2012, SIAM J. Sci. Comput..

[4]  E Weinan,et al.  The heterogeneous multiscale method* , 2012, Acta Numerica.

[5]  E. Weinan.  Principles of Multiscale Modeling , 2011 .

[6]  Jerrold E. Marsden,et al.  Nonintrusive and Structure Preserving Multiscale Integration of Stiff ODEs, SDEs, and Hamiltonian Systems with Hidden Slow Dynamics via Flow Averaging , 2009, Multiscale Model. Simul..

[7]  E Weinan,et al.  A general strategy for designing seamless multiscale methods , 2009, J. Comput. Phys..

[8]  Grigorios A. Pavliotis,et al.  Multiscale Methods: Averaging and Homogenization , 2008 .

[9]  Eric Vanden-Eijnden,et al.  ON HMM-like integrators and projective integration methods for systems with multiple time scales , 2007 .

[10]  E Weinan,et al.  Heterogeneous multiscale methods: A review , 2007 .

[11]  Ioannis G. Kevrekidis,et al.  Strong convergence of projective integration schemes for singularly perturbed stochastic differential systems , 2006 .

[12]  E. Vanden-Eijnden,et al.  Analysis of multiscale methods for stochastic differential equations , 2005 .

[13]  C. Hwang,et al.  Accelerating diffusions , 2005, math/0505245.

[14]  Mark Freidlin,et al.  Random perturbations of dynamical systems and diffusion processes with conservation laws , 2004 .

[15]  Tianhai Tian,et al.  Stiffly accurate Runge-Kutta methods for stiff stochastic differential equations , 2001 .

[16]  Mark Freidlin,et al.  On stochastic behavior of perturbed Hamiltonian systems , 2000, Ergodic Theory and Dynamical Systems.

[17]  R. Tweedie,et al.  Exponential convergence of Langevin distributions and their discrete approximations , 1996 .

[18]  Mark Freidlin,et al.  Random perturbations of Hamiltonian systems , 1994 .

[19]  Mark Freidlin,et al.  Diffusion Processes on Graphs and the Averaging Principle , 1993 .

[20]  J. Banavar,et al.  The Heterogeneous Multi-Scale Method , 1992 .

[21]  A. Skorokhod.  Asymptotic Methods in the Theory of Stochastic Differential Equations , 2008 .

[22]  M. Freidlin,et al.  Random Perturbations of Dynamical Systems , 1984 .