Optimal scalings for local Metropolis--Hastings chains on nonproduct targets in high dimensions
arXiv:0908.0865 · doi:10.1214/08-AAP563
Abstract
We investigate local MCMC algorithms, namely the random-walk Metropolis and the Langevin algorithms, and identify the optimal choice of the local step-size as a function of the dimension of the state space, asymptotically as . We consider target distributions defined as a change of measure from a product law. Such structures arise, for instance, in inverse problems or Bayesian contexts when a product prior is combined with the likelihood. We state analytical results on the asymptotic behavior of the algorithms under general conditions on the change of measure. Our theory is motivated by applications on conditioned diffusion processes and inverse problems related to the 2D Navier--Stokes equation.
Published in at http://dx.doi.org/10.1214/08-AAP563 the Annals of Applied Probability (http://www.imstat.org/aap/) by the Institute of Mathematical Statistics (http://www.imstat.org)
References in corpus (4)
Cited by in corpus (6)
- A Survey of Monte Carlo Methods for Parameter Estimation
- Bayesian computation: a perspective on the current state, and sampling backwards and forwards
- Self-Tuning Hamiltonian Monte Carlo for Accelerated Sampling
- Efficiency and computability of MCMC with Langevin, Hamiltonian, and other matrix-splitting proposals
- Scaling Up Bayesian Uncertainty Quantification for Inverse Problems using Deep Neural Networks
- Sequential Monte Carlo Methods for Bayesian Elliptic Inverse Problems