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

Download:

Current browse context:

stat.ME

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 > Methodology

Title: Optimal Thinning of MCMC Output

Abstract: The use of heuristics to assess the convergence and compress the output of Markov chain Monte Carlo can be sub-optimal in terms of the empirical approximations that are produced. Typically a number of the initial states are attributed to "burn in" and removed, whilst the remainder of the chain is "thinned" if compression is also required. In this paper we consider the problem of retrospectively selecting a subset of states, of fixed cardinality, from the sample path such that the approximation provided by their empirical distribution is close to optimal. A novel method is proposed, based on greedy minimisation of a kernel Stein discrepancy, that is suitable for problems where heavy compression is required. Theoretical results guarantee consistency of the method and its effectiveness is demonstrated in the challenging context of parameter inference for ordinary differential equations. Software is available in the Stein Thinning package in Python, R and MATLAB.
Comments: To appear in the Journal of the Royal Statistical Society, Series B, 2022+
Subjects: Methodology (stat.ME); Statistics Theory (math.ST); Computation (stat.CO); Machine Learning (stat.ML)
Cite as: arXiv:2005.03952 [stat.ME]
  (or arXiv:2005.03952v5 [stat.ME] for this version)

Submission history

From: Chris Oates [view email]
[v1] Fri, 8 May 2020 10:54:25 GMT (5226kb,D)
[v2] Mon, 29 Jun 2020 18:48:04 GMT (9091kb,D)
[v3] Mon, 12 Jul 2021 14:58:49 GMT (25762kb,D)
[v4] Mon, 20 Sep 2021 09:16:46 GMT (25762kb,D)
[v5] Tue, 11 Jan 2022 11:21:12 GMT (25762kb,D)

Link back to: arXiv, form interface, contact.