On the geometric ergodicity of Hamiltonian Monte Carlo
From MaRDI portal
Publication:2325354
Abstract: We establish general conditions under which Markov chains produced by the Hamiltonian Monte Carlo method will and will not be geometrically ergodic. We consider implementations with both position-independent and position-dependent integration times. In the former case we find that the conditions for geometric ergodicity are essentially a gradient of the log-density which asymptotically points towards the centre of the space and grows no faster than linearly. In an idealised scenario in which the integration time is allowed to change in different regions of the space, we show that geometric ergodicity can be recovered for a much broader class of tail behaviours, leading to some guidelines for the choice of this free parameter in practice.
Recommendations
Cites work
- scientific article; zbMATH DE number 3066379 (Why is no real title available?)
- A function space HMC algorithm with second order Langevin diffusion limit
- A general metric for Riemannian manifold Hamiltonian Monte Carlo
- A note on Metropolis-Hastings kernels for general state spaces
- An introduction to MCMC for machine learning
- Connections and Extensions: A Discussion of the Paper by Girolami and Byrne
- Equation of state calculations by fast computing machines
- Ergodicity for SDEs and approximations: locally Lipschitz vector fields and degenerate noise.
- Error bounds for Metropolis-Hastings algorithms applied to perturbations of Gaussian measures in high dimensions
- Exponential convergence of Langevin distributions and their discrete approximations
- Extra chance generalized hybrid Monte Carlo
- General state space Markov chains and MCMC algorithms
- Geodesic Monte Carlo on embedded manifolds
- Geometric convergence and central limit theorems for multidimensional Hastings and Metropolis algorithms
- Geometric ergodicity and hybrid Markov chains
- Geometric ergodicity of Metropolis algorithms
- Geometric integrators and the Hamiltonian Monte Carlo method
- Honest exploration of intractable probability distributions via Markov chain Monte Carlo.
- Hybrid Monte Carlo on Hilbert spaces
- Information-geometric Markov chain Monte Carlo methods using diffusions
- Inverse problems: a Bayesian perspective
- Irreducibility and geometric ergodicity of Hamiltonian Monte Carlo
- Iterated Random Functions
- MCMC using Hamiltonian dynamics
- Markov chains and stochastic stability
- Markov chains for exploring posterior distributions. (With discussion)
- Monte Carlo sampling methods using Markov chains and their applications
- Necessary conditions for geometric and polynomial ergodicity of random-walk-type Markov chains
- Nonasymptotic mixing of the MALA algorithm
- On the geometric ergodicity of Hamiltonian Monte Carlo
- Optimal tuning of the hybrid Monte Carlo algorithm
- Quantitative bounds of convergence for geometrically ergodic Markov chain in the Wasserstein distance with application to the Metropolis adjusted Langevin algorithm
- Randomized Hamiltonian Monte Carlo
- Ricci curvature of Markov chains on metric spaces
- Riemann manifold Langevin and Hamiltonian Monte Carlo methods. With discussion and authors' reply
- Simulating Hamiltonian Dynamics
- Some things we've learned (about Markov chain Monte Carlo)
- The geometric foundations of Hamiltonian Monte Carlo
- The no-U-turn sampler: adaptively setting path lengths in Hamiltonian Monte Carlo
- Theoretical and numerical comparison of some sampling methods for molecular dynamics
Cited in
(39)- \(\pi\) VAE: a stochastic process prior for Bayesian deep learning with MCMC
- Fast Markov Chain Monte Carlo for High-Dimensional Bayesian Regression Models With Shrinkage Priors
- Speed up Zig-Zag
- Irreducibility and geometric ergodicity of Hamiltonian Monte Carlo
- On the geometric ergodicity of Hamiltonian Monte Carlo
- Stochastic approximation Hamiltonian Monte Carlo
- On the geometric ergodicity of Metropolis-Hastings algorithms
- Monte Carlo Simulation on the Stiefel Manifold via Polar Expansion
- Randomized Hamiltonian Monte Carlo as scaling limit of the bouncy particle sampler and dimension-free convergence rates
- An entropic approach for Hamiltonian Monte Carlo: the idealized case
- Two-scale coupling for preconditioned Hamiltonian Monte Carlo in infinite dimensions
- Geometric Ergodicity for Hamiltonian Monte Carlo on Compact Manifolds
- V-subgeometric ergodicity for a Hastings-Metropolis algorithm
- Geometric ergodicity of Gibbs samplers for the horseshoe and its regularized variants
- Mixing of Metropolis-adjusted Markov chains via couplings: the high acceptance regime
- Learning variational autoencoders via MCMC speed measures
- Modified Hamiltonian Monte Carlo for Bayesian inference
- Efficient shape-constrained inference for the autocovariance sequence from a reversible Markov chain
- Couplings for Andersen dynamics
- Randomized Hamiltonian Monte Carlo
- Variance bounding of delayed-acceptance kernels
- On \(L^\mathfrak{q}\) convergence of the Hamiltonian Monte Carlo
- On the Geometric Ergodicity of Metropolis-Hastings Algorithms for Lattice Gaussian Sampling
- Convergence of unadjusted Hamiltonian Monte Carlo for mean-field models
- On the geometric ergodicity of hybrid samplers
- Mixing rates for Hamiltonian Monte Carlo algorithms in finite and infinite dimensions
- Mixing of Hamiltonian Monte Carlo on strongly log-concave distributions: continuous dynamics
- Adaptive weighting of Bayesian physics informed neural networks for multitask and multiscale forward and inverse problems
- Recycling intermediate steps to improve Hamiltonian Monte Carlo
- Convergence rates of Metropolis-Hastings algorithms
- Convergence of Position-Dependent MALA with Application to Conditional Simulation in GLMMs
- Kinetic energy choice in Hamiltonian/hybrid Monte Carlo
- Emulation-accelerated Hamiltonian Monte Carlo algorithms for parameter estimation and uncertainty quantification in differential equation models
- The geometric foundations of Hamiltonian Monte Carlo
- Geometric integrators and the Hamiltonian Monte Carlo method
- Approximations of geometrically ergodic reversible Markov chains
- On \(L^2\) convergence of the Hamiltonian Monte Carlo
- Bayesian modelling of time-varying conditional heteroscedasticity
- HMC: reducing the number of rejections by not using leapfrog and some results on the acceptance rate
This page was built for publication: On the geometric ergodicity of Hamiltonian Monte Carlo
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2325354)
