Optimal Scaling and Diffusion Limits for the Langevin Algorithm in High Dimensions

The Metropolis-adjusted Langevin (MALA) algorithm is a sampling algorithm which makes local moves by incorporating information about the gradient of the logarithm of the target density. In this paper we study the efficiency of MALA on a natural class of target measures supported on an infinite dimensional Hilbert space. These natural measures have density with respect to a Gaussian random field measure and arise in many applications such as Bayesian nonparametric statistics and the theory of conditioned diffusions. We prove that, started in stationarity, a suitably interpolated and scaled version of the Markov chain corresponding to MALA converges to an infinite dimensional diffusion process. Our results imply that, in stationarity, the MALA algorithm applied to an N-dimensional approximation of the target will take $\mathcal{O}(N^{1/3})$ steps to explore the invariant measure, comparing favorably with the Random Walk Metropolis which was recently shown to require $\mathcal{O}(N)$ steps when applied to the same class of problems.

[1]  N. Metropolis,et al.  Equation of State Calculations by Fast Computing Machines , 1953, Resonance.

[2]  E. Berger.  Asymptotic behaviour of a class of stochastic approximation procedures , 1986 .

[3]  R. Durrett.  Stochastic Calculus: A Practical Introduction , 1996 .

[4]  A. Gelman,et al.  Weak convergence and optimal scaling of random walk Metropolis algorithms , 1997 .

[5]  J. Rosenthal,et al.  Optimal scaling of discrete approximations to Langevin diffusions , 1998 .

[6]  Hoon Kim,et al.  Monte Carlo Statistical Methods , 2000, Technometrics.

[7]  Gareth O. Roberts,et al.  From metropolis to diffusions: Gibbs states and optimal scaling , 2000 .

[8]  J. Rosenthal,et al.  Optimal scaling for various Metropolis-Hastings algorithms , 2001 .

[9]  Sergio Scarlatti,et al.  Optimal scaling of MaLa for nonlinear regression , 2004 .

[10]  J. Rosenthal,et al.  Scaling limits for the transient phase of local Metropolis–Hastings algorithms , 2005 .

[11]  Christian P. Robert,et al.  Monte Carlo Statistical Methods (Springer Texts in Statistics) , 2005 .

[12]  J. Voss,et al.  Analysis of SPDEs arising in path sampling. Part I: The Gaussian case , 2005 .

[13]  S. Ethier,et al.  Markov Processes: Characterization and Convergence , 2005 .

[14]  T. Kurtz,et al.  Stochastic equations in infinite dimensions , 2006 .

[15]  A. Stuart,et al.  ANALYSIS OF SPDES ARISING IN PATH SAMPLING PART II: THE NONLINEAR CASE , 2006, math/0601092.

[16]  M. B'edard.  Weak convergence of Metropolis algorithms for non-i.i.d. target distributions , 2007, 0710.3684.

[17]  G. Roberts,et al.  An MCMC method for diffusion bridges , 2008 .

[18]  G. Roberts,et al.  MCMC methods for diffusion bridges , 2008 .

[19]  Jonathan C. Mattingly,et al.  SPDE limits of the random walk Metropolis algorithm in high dimensions , 2009 .

[20]  Matti Lassas. Eero Saksman,et al.  Discretization-invariant Bayesian inversion and Besov space priors , 2009, 0901.4220.

[21]  A. Stuart,et al.  MCMC methods for sampling function space , 2009 .

[22]  Gareth Roberts,et al.  Optimal scalings for local Metropolis--Hastings chains on nonproduct targets in high dimensions , 2009, 0908.0865.

[23]  J. M. Sanz-Serna,et al.  Optimal tuning of the hybrid Monte Carlo algorithm , 2010, 1001.4460.

[24]  Andrew M. Stuart,et al.  Inverse problems: A Bayesian perspective , 2010, Acta Numerica.

[25]  P. Fearnhead,et al.  The Random Walk Metropolis: Linking Theory and Practice Through a Case Study , 2010, 1011.6217.

[26]  A. Stuart,et al.  Besov priors for Bayesian inverse problems , 2011, 1105.0889.

[27]  J. M. Sanz-Serna,et al.  Hybrid Monte Carlo on Hilbert spaces , 2011 .

[28]  A. Stuart,et al.  Signal processing problems on function space: Bayesian formulation, stochastic PDEs and effective MCMC methods , 2011 .

[29]  A. M. Stuart,et al.  Sparse deterministic approximation of Bayesian inverse problems , 2011, 1103.4522.

[30]  Jonathan C. Mattingly,et al.  Diffusion limits of the random walk metropolis algorithm in high dimensions , 2010, 1003.4306.