Stability of partially implicit Langevin schemes and their MCMC variants

DOI10.1007/S11009-010-9196-5MaRDI QIDQ429995zbMATH OpenOpenAlexFDO

Authors Bruno Casella, O. Stramer, Gareth O. Roberts

Publication date 20 June 2012

Published in Methodology and Computing in Applied Probability (Search for Journal in Brave)

Full work available at URL https://mpra.ub.uni-muenchen.de/95220/1/MPRA_paper_95220.pdf

zbMATH Keywords

ergodicityLangevin equationMetropolis-Hastings algorithmpartially implicit methodstime discretization

Mathematics Subject Classification ID

Computational methods for stochastic equations (aspects of stochastic analysis) (60H35) Computational methods in Markov chains (60J22) Monte Carlo methods (65C05) Numerical solutions to stochastic differential and integral equations (65C30)

Review text

Ergodicity properties of implicit and partially implicit discretization schemes for Langevin equations corresponding to a class of desired stationary distributions are established. It is shown that when used as a proposal for a Metropolis-Hastings scheme, the ergodicity properties are retained. A partially implicit scheme based on local linearization is also analyzed. Examples are provided. For ease of exposition, only the scalar case is considered.

Recommendations

• Langevin diffusions and Metropolis-Hastings algorithms
• Langevin-type models. I: Diffusions with given stationary distributions and their discretizations
• Weak backward error analysis for Langevin process
• Langevin diffusions and the Metropolis-adjusted Langevin algorithm
• Optimal Scaling of Discrete Approximations to Langevin Diffusions

Cites work

• A factorisation of diffusion measure and finite sample path constructions
• Ergodicity for SDEs and approximations: locally Lipschitz vector fields and degenerate noise.
• Estimation for nonlinear stochastic differential equations by a local linearization method¹
• Exponential convergence of Langevin distributions and their discrete approximations
• Geometric convergence and central limit theorems for multidimensional Hastings and Metropolis algorithms
• Geometric ergodicity of discrete-time approximations to multivariate diffusions
• Higher-order implicit strong numerical schemes for stochastic differential equations
• scientific article; zbMATH DE number 979834 (Why is no real title available?)
• Langevin-type models II: Self-targeting candidates for MCMC algorithms
• MCMC METHODS FOR DIFFUSION BRIDGES
• Mean-Square and Asymptotic Stability of the Stochastic Theta Method
• Rates of convergence of the Hastings and Metropolis algorithms
• Stability Analysis of Numerical Schemes for Stochastic Differential Equations

Cited in

(8)

• Langevin diffusions and Metropolis-Hastings algorithms
• Proximal Markov chain Monte Carlo algorithms
• A patch that imparts unconditional stability to explicit integrators for Langevin-like equations
• scientific article; zbMATH DE number 7415079 (Why is no real title available?)
• Weak backward error analysis for overdamped Langevin processes
• Transport map unadjusted Langevin algorithms: learning and discretizing perturbed samplers
• Weak backward error analysis for Langevin process
• Stability of sequential Monte Carlo samplers via the Foster-Lyapunov condition

This page was built for publication: Stability of partially implicit Langevin schemes and their MCMC variants

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q429995)