A hybrid adaptive MCMC algorithm in function spaces
From MaRDI portal
Abstract: The preconditioned Crank-Nicolson (pCN) method is a Markov Chain Monte Carlo (MCMC) scheme, specifically designed to perform Bayesian inferences in function spaces. Unlike many standard MCMC algorithms, the pCN method can preserve the sampling efficiency under the mesh refinement, a property referred to as being dimension independent. In this work we consider an adaptive strategy to further improve the efficiency of pCN. In particular we develop a hybrid adaptive MCMC method: the algorithm performs an adaptive Metropolis scheme in a chosen finite dimensional subspace, and a standard pCN algorithm in the complement space of the chosen subspace. We show that the proposed algorithm satisfies certain important ergodicity conditions. Finally with numerical examples we demonstrate that the proposed method has competitive performance with existing adaptive algorithms.
Recommendations
- On an adaptive preconditioned Crank-Nicolson MCMC algorithm for infinite dimensional Bayesian inference
- On a generalization of the preconditioned Crank-Nicolson metropolis algorithm
- MCMC methods for functions: modifying old algorithms to make them faster
- A function space HMC algorithm with second order Langevin diffusion limit
- A numerical method for solving linear systems in the preconditioned Crank-Nicolson algorithm
Cites work
- A TV-Gaussian prior for infinite-dimensional Bayesian inverse problems and its numerical implementations
- Accelerated dimension-independent adaptive metropolis
- Adaptive construction of surrogates for the Bayesian solution of inverse problems
- An adaptive Metropolis algorithm
- An introduction to infinite-dimensional analysis
- Coupling and Ergodicity of Adaptive Markov Chain Monte Carlo Algorithms
- Data-driven model reduction for the Bayesian solution of inverse problems
- Diffusion limits of the random walk Metropolis algorithm in high dimensions
- Dimension-independent likelihood-informed MCMC
- Gaussian processes for machine learning.
- scientific article; zbMATH DE number 2117879 (Why is no real title available?)
- Inverse problems: a Bayesian perspective
- MCMC methods for functions: modifying old algorithms to make them faster
- On a generalization of the preconditioned Crank-Nicolson metropolis algorithm
- On an adaptive preconditioned Crank-Nicolson MCMC algorithm for infinite dimensional Bayesian inference
- On the containment condition for adaptive Markov chain Monte Carlo algorithms
- Optimal scaling for various Metropolis-Hastings algorithms.
- Proposals which speed up function-space MCMC
- Statistical and computational inverse problems.
- Stochastic collocation algorithms using \(l_1\)-minimization for Bayesian solution of inverse problems
- The identification of a Robin coefficient by a conjugate gradient method
Cited in
(7)- On a generalization of the preconditioned Crank-Nicolson metropolis algorithm
- Ensemble sampler for infinite-dimensional inverse problems
- A numerical method for solving linear systems in the preconditioned Crank-Nicolson algorithm
- An Adaptive Independence Sampler MCMC Algorithm for Bayesian Inferences of Functions
- Error control of the numerical posterior with Bayes factors in Bayesian uncertainty quantification
- Multilevel Delayed Acceptance MCMC
- An extended sampling-Bayesian approach for inverse elastic scattering problems with limited aperture data
This page was built for publication: A hybrid adaptive MCMC algorithm in function spaces
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4636400)
