Multilevel Delayed Acceptance MCMC
From MaRDI portal
Abstract: We develop a novel Markov chain Monte Carlo (MCMC) method that exploits a hierarchy of models of increasing complexity to efficiently generate samples from an unnormalized target distribution. Broadly, the method rewrites the Multilevel MCMC approach of Dodwell et al. (2015) in terms of the Delayed Acceptance (DA) MCMC of Christen & Fox (2005). In particular, DA is extended to use a hierarchy of models of arbitrary depth, and allow subchains of arbitrary length. We show that the algorithm satisfies detailed balance, hence is ergodic for the target distribution. Furthermore, multilevel variance reduction is derived that exploits the multiple levels and subchains, and an adaptive multilevel correction to coarse-level biases is developed. Three numerical examples of Bayesian inverse problems are presented that demonstrate the advantages of these novel methods. The software and examples are available in PyMC3.
Recommendations
- Multilevel Markov Chain Monte Carlo
- Analysis of a Class of Multilevel Markov Chain Monte Carlo Algorithms Based on Independent Metropolis–Hastings
- Efficiency of delayed-acceptance random walk metropolis algorithms
- Multilevel sequential Monte Carlo samplers
- Multilevel Monte Carlo in approximate Bayesian computation
Cites work
- A Hierarchical Multilevel Markov Chain Monte Carlo Algorithm with Applications to Uncertainty Quantification in Subsurface Flow
- A hybrid adaptive MCMC algorithm in function spaces
- A MULTI-INDEX MARKOV CHAIN MONTE CARLO METHOD
- A short history of mathematical population dynamics.
- Accelerating Markov chain Monte Carlo with active subspaces
- Accelerating uncertainty quantification of groundwater flow modelling using a deep neural network proxy
- Adaptive dimension reduction to accelerate infinite-dimensional geometric Markov chain Monte Carlo
- An adaptive Metropolis algorithm
- An adaptive version for the Metropolis adjusted Langevin algorithm with a truncated drift
- Coupling and Ergodicity of Adaptive Markov Chain Monte Carlo Algorithms
- Dimensionality reduction and polynomial chaos acceleration of Bayesian inference in inverse problems
- Discrete inverse problems. Insight and algorithms.
- Equation of state calculations by fast computing machines
- Exponential convergence of Langevin distributions and their discrete approximations
- Finite element error analysis of elliptic PDEs with random coefficients and its application to multilevel Monte Carlo methods
- Further analysis of multilevel Monte Carlo methods for elliptic PDEs with random coefficients
- Gaussian processes for machine learning.
- General state space Markov chains and MCMC algorithms
- Geometric MCMC for infinite-dimensional inverse problems
- scientific article; zbMATH DE number 3764843 (Why is no real title available?)
- scientific article; zbMATH DE number 2000348 (Why is no real title available?)
- Monte Carlo sampling methods using Markov chains and their applications
- Multi-index stochastic collocation convergence rates for random PDEs with parametric regularity
- Multi-level Monte Carlo finite element method for elliptic PDEs with stochastic coefficients
- Multilevel Monte Carlo methods and applications to elliptic PDEs with random coefficients
- Multilevel Monte Carlo Path Simulation
- Optimal Scaling of Discrete Approximations to Langevin Diffusions
- Parallel local approximation MCMC for expensive models
- Randomized reduced forward models for efficient Metropolis-Hastings MCMC, with application to subsurface fluid flow and capacitance tomography
- Rank-normalization, folding, and localization: an improved \(\widehat{R}\) for assessing convergence of MCMC (with Discussion)
- Statistical inverse problems: discretization, model reduction and inverse crimes
- Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images
- Stochastic spectral methods for efficient Bayesian solution of inverse problems
- The Multiple-Try Method and Local Optimization in Metropolis Sampling
- The no-U-turn sampler: adaptively setting path lengths in Hamiltonian Monte Carlo
Cited in
(15)- Variance bounding of delayed-acceptance kernels
- Multilevel Markov Chain Monte Carlo
- Efficient Marginalization-Based MCMC Methods for Hierarchical Bayesian Inverse Problems
- Analysis of a Class of Multilevel Markov Chain Monte Carlo Algorithms Based on Independent Metropolis–Hastings
- Multilevel dimension-independent likelihood-informed MCMC for large-scale inverse problems
- A posteriori stochastic correction of reduced models in delayed-acceptance MCMC, with application to multiphase subsurface inverse problems
- Sequential Kalman tuning of the t-preconditioned Crank-Nicolson algorithm: efficient, adaptive and gradient-free inference for Bayesian inverse problems
- Democratizing uncertainty quantification
- A hybrid two-level MCMC framework to accelerate posterior mean estimation with deep learning surrogates for Bayesian inverse problems
- An implicit-explicit Monte Carlo method for semi-linear PDEs driven by -stable Lévy process and its error estimates
- Derivative-informed neural operator acceleration of geometric MCMC for infinite-dimensional Bayesian inverse problems
- Multilevel Markov chain Monte Carlo with likelihood scaling for Bayesian inversion with high-resolution observations
- Local sensitivity analysis for Bayesian inverse problems
- Fused ensembles of dynamic-rupture earthquake simulations to accelerate Bayesian inference
- A localized consensus-based sampling algorithm
This page was built for publication: Multilevel Delayed Acceptance MCMC
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6164127)
