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


Current browse context:


Change to browse by:

References & Citations


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

Statistics > Methodology

Title: Posterior Integration on a Riemannian Manifold

Abstract: The geodesic Markov chain Monte Carlo method and its variants enable computation of integrals with respect to a posterior supported on a manifold. However, for regular integrals, the convergence rate of the ergodic average will be sub-optimal. To fill this gap, this paper extends the efficient posterior integration method of Oates et al. (2017) to the case of a Riemannian manifold. In contrast to the original Euclidean case, no non-trivial boundary conditions are needed for a closed manifold. The method is assessed through simulation and deployed to compute posterior integrals for an Australian Mesozoic paleomagnetic pole model, whose parameters are constrained to lie on the manifold $M = \mathbb{S}^2 \times \mathbb{R}_+$.
Comments: This paper was superseded by arXiv:1810.04946
Subjects: Methodology (stat.ME)
Cite as: arXiv:1712.01793 [stat.ME]
  (or arXiv:1712.01793v4 [stat.ME] for this version)

Submission history

From: Chris Oates [view email]
[v1] Tue, 5 Dec 2017 18:15:39 GMT (1939kb,D)
[v2] Fri, 12 Jan 2018 18:13:43 GMT (779kb,D)
[v3] Mon, 15 Jan 2018 10:13:25 GMT (781kb,D)
[v4] Sun, 14 Oct 2018 11:26:27 GMT (781kb,D)

Link back to: arXiv, form interface, contact.