On the convergence of Hamiltonian Monte Carlo

2017/04/29 by Alain Durmus, Éric Moulines, Durmus, Alain +3 · 5 citations
Mathematics · #Computation (stat.CO) #FOS: Computer and information sciences #FOS: Mathematics #Markov Chains and Monte Carlo Methods #Mathematical Approximation and Integration #Probability (math.PR) #Stochastic processes and statistical mechanics

paper · pdf · doi:10.48550/arxiv.1705.00166

openalex publication_date 2017/04/29 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

This paper discusses the irreducibility and geometric ergodicity of the Hamiltonian Monte Carlo (HMC) algorithm. We consider cases where the number of steps of the symplectic integrator is either fixed or random. Under mild conditions on the potential \F associated with target distribution π, we first show that the Markov kernel associated to the HMC algorithm is irreducible and recurrent. Under more stringent conditions, we then establish that the Markov kernel is Harris recurrent. Finally, we provide verifiable conditions on \F under which the HMC sampler is geometrically ergodic.

Citations

• Learning Generative Models with Visual Attention

Cited by

• On the convergence of dynamic implementations of Hamiltonian Monte Carlo and no U-turn samplers
• Entropy-based adaptive Hamiltonian Monte Carlo
• Fast Convergence of Langevin Dynamics on Manifold: Geodesics meet Log-Sobolev
• Nonreversible MCMC from conditional invertible transforms: a complete\n recipe with convergence guarantees
• An Introduction to Hamiltonian Monte Carlo Method for Sampling

Related

• Mixing Rates for Hamiltonian Monte Carlo Algorithms in Finite and Infinite Dimensions
• On the Geometric Ergodicity of Hamiltonian Monte Carlo
• On the convergence of dynamic implementations of Hamiltonian Monte Carlo and no U-turn samplers
• On Numerical Considerations for Riemannian Manifold Hamiltonian Monte Carlo
• An Introduction to Hamiltonian Monte Carlo Method for Sampling