Home > Research > Publications & Outputs > Reversible Jump PDMP Samplers for Variable Sele...

Electronic data

  • RJ_PDMP_Samplers

    Accepted author manuscript, 7.55 MB, PDF document

    Available under license: CC BY: Creative Commons Attribution 4.0 International License

  • 2010.11771v1

    Other version, 815 KB, PDF document

Keywords

View graph of relations

Reversible Jump PDMP Samplers for Variable Selection

Research output: Contribution to Journal/MagazineJournal articlepeer-review

E-pub ahead of print
<mark>Journal publication date</mark>11/07/2022
<mark>Journal</mark>Journal of the American Statistical Association
Publication StatusE-pub ahead of print
Early online date11/07/22
<mark>Original language</mark>Undefined/Unknown

Abstract

A new class of Markov chain Monte Carlo (MCMC) algorithms, based on simulating piecewise deterministic Markov processes (PDMPs), have recently shown great promise: they are non-reversible, can mix better than standard MCMC algorithms, and can use subsampling ideas to speed up computation in big data scenarios. However, current PDMP samplers can only sample from posterior densities that are differentiable almost everywhere, which precludes their use for model choice. Motivated by variable selection problems, we show how to develop reversible jump PDMP samplers that can jointly explore the discrete space of models and the continuous space of parameters. Our framework is general: it takes any existing PDMP sampler, and adds two types of trans-dimensional moves that allow for the addition or removal of a variable from the model. We show how the rates of these trans-dimensional moves can be calculated so that the sampler has the correct invariant distribution. Simulations show that the new samplers can mix better than standard MCMC algorithms. Our empirical results show they are also more efficient than gradient-based samplers that avoid model choice through use of continuous spike-and-slab priors which replace a point mass at zero for each parameter with a density concentrated around zero.

Bibliographic note

Code available from https://github.com/matt-sutton/rjpdmp