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: Bayesian high-dimensional covariate selection in non-linear mixed-effects models using the SAEM algorithm

Authors: Maud Delattre (MaIAGE), Guillaume Kon Kam King (MaIAGE), Marion Naveau (MIA Paris-Saclay, MaIAGE), Laure Sansonnet (MIA Paris-Saclay)
Abstract: High-dimensional data, with many more covariates than observations, such as genomic data for example, are now commonly analysed. In this context, it is often desirable to be able to focus on the few most relevant covariates through a variable selection procedure. High-dimensional variable selection is widely documented in standard regression models, but there are still few tools to address it in the context of non-linear mixed-effects models. In this work, variable selection is approached from a Bayesian perspective and a selection procedure is proposed, combining the use of spike-and-slab priors and the SAEM algorithm. Similarly to LASSO regression, the set of relevant covariates is selected by exploring a grid of values for the penalisation parameter. The proposed approach is much faster than a classical MCMC algorithm and shows very good selection performances on simulated data.
Subjects: Statistics Theory (math.ST)
Cite as: arXiv:2206.01012 [math.ST]
  (or arXiv:2206.01012v1 [math.ST] for this version)

Submission history

From: Marion Naveau [view email]
[v1] Thu, 2 Jun 2022 12:22:12 GMT (1257kb,D)
[v2] Mon, 21 Nov 2022 13:44:46 GMT (1172kb,D)

Link back to: arXiv, form interface, contact.