EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Does Hamiltonian Monte Carlo mix faster than a random walk on multimodal densities?

Oren Mangoubi, Natesh S. Pillai, Aaron Smith

Advances in Applied Probability · 23.04.2026 · 1-48

DOI: 10.1017/apr.2026.10061 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Advances in Applied Probability, online erschienen am 23.04.2026. Verfasst von Oren Mangoubi und 2 weiteren Autor:innen. Die Arbeit nennt 61 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Hamiltonian Monte Carlo (HMC) is a very popular collection of Markov chain Monte Carlo (MCMC) algorithms. One explanation for the popularity of HMC algorithms is their excellent performance as the dimension d of the target becomes large: theoretical analyses show that popular versions of HMC can have a running time that scales as well as $d^{0.25}$ in good conditions, while even an optimally tuned random-walk metropolis (RWM) algorithm will not do better than d . In this paper, we investigate a different scaling question: does HMC beat RWM for targets with well-separated modes? We find that the answer is often no . Our main tool for answering this question is a novel and simple formula for the conductance of HMC based on Liouville’s theorem, and we also show how this new formula can be used to give very short proofs of results that seem tedious to show with the usual formula. We also use this result to compute the spectral gap of HMC algorithms, for both the classical HMC with isotropic momentum and the recent Riemannian HMC, for multimodal targets. While we focus on the concrete comparison of RWM and HMC, we expect qualitatively similar conclusions to hold for other gradient-based algorithms.

Zitieren

APA

Mangoubi, O., Pillai, N. S., & Smith, A. (2026). Does Hamiltonian Monte Carlo mix faster than a random walk on multimodal densities?. Advances in Applied Probability, 1-48. https://doi.org/10.1017/apr.2026.10061

Vancouver

Mangoubi O, Pillai NS, Smith A. Does Hamiltonian Monte Carlo mix faster than a random walk on multimodal densities?. Advances in Applied Probability. 2026:1-48. doi:10.1017/apr.2026.10061

BibTeX
@article{Mangoubi2026,
  author = {Mangoubi, Oren and Pillai, Natesh S. and Smith, Aaron},
  title = {Does Hamiltonian Monte Carlo mix faster than a random walk on multimodal densities?},
  journal = {Advances in Applied Probability},
  year = {2026},
  pages = {1--48},
  doi = {10.1017/apr.2026.10061}
}
Kostenlos starten →