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

Download:

Current browse context:

math.ST

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

Mathematics > Statistics Theory

Title: Stick-breaking processes with exchangeable length variables

Abstract: Our object of study is the general class of stick-breaking processes with exchangeable length variables. These generalize well-known Bayesian non-parametric priors in an unexplored direction. We give conditions to assure the respective species sampling process is proper and the corresponding prior has full support. For a rich sub-class we explain how, by tuning a single $[0,1]$-valued parameter, the stochastic ordering of the weights can be modulated, and Dirichlet and Geometric priors can be recovered. A general formula for the distribution of the latent allocation variables is derived and an MCMC algorithm is proposed for density estimation purposes.
Comments: Accepted for publication by the Journal of the American Statistical Association. 44 pages, 11 figures, supplementary material
Subjects: Statistics Theory (math.ST); Probability (math.PR); Methodology (stat.ME); Other Statistics (stat.OT)
Cite as: arXiv:2008.04475 [math.ST]
  (or arXiv:2008.04475v2 [math.ST] for this version)

Submission history

From: María Fernanda Gil-Leyva Villa [view email]
[v1] Tue, 11 Aug 2020 02:00:04 GMT (306kb,D)
[v2] Sun, 18 Jul 2021 07:09:07 GMT (4491kb,D)

Link back to: arXiv, form interface, contact.