We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.CO

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Computation

Title: Boost your favorite Markov Chain Monte Carlo sampler using Kac's theorem: the Kick-Kac teleportation algorithm

Abstract: The present paper focuses on the problem of sampling from a given target distribution $\pi$ defined on some general state space. To this end, we introduce a novel class of non-reversible Markov chains, each chain being defined on an extended state space and having an invariant probability measure admitting $\pi$ as a marginal distribution. The proposed methodology is inspired by a new formulation of Kac's theorem and allows global and local dynamics to be smoothly combined. Under mild conditions, the corresponding Markov transition kernel can be shown to be irreducible and Harris recurrent. In addition, we establish that geometric ergodicity holds under appropriate conditions on the global and local dynamics. Finally, we illustrate numerically the use of the proposed method and its potential benefits in comparison to existing Markov chain Monte Carlo (MCMC) algorithms.
Subjects: Computation (stat.CO); Numerical Analysis (math.NA); Probability (math.PR)
MSC classes: 62-08, 60J20, 65C05 (Primary), 60F15 (Secondary)
Cite as: arXiv:2201.05002 [stat.CO]
  (or arXiv:2201.05002v1 [stat.CO] for this version)

Submission history

From: Aurélien Enfroy [view email]
[v1] Thu, 13 Jan 2022 14:33:57 GMT (997kb,D)

Link back to: arXiv, form interface, contact.