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: Spike-and-Slab Group Lasso for Consistent Estimation and Variable Selection in Non-Gaussian Generalized Additive Models

Authors: Ray Bai
Abstract: We study estimation and variable selection in non-Gaussian Bayesian generalized additive models (GAMs) under a spike-and-slab prior for grouped variables. Our framework subsumes GAMs for logistic regression, Poisson regression, negative binomial regression, and gamma regression, and encompasses both canonical and non-canonical link functions. Under mild conditions, we establish posterior contraction rates and model selection consistency when $p \gg n$. For computation, we propose an EM algorithm for obtaining MAP estimates in our model, which is available in the R package sparseGAM. We illustrate our method on both synthetic and real data sets.
Comments: 35 pages, 4 figures
Subjects: Methodology (stat.ME)
Cite as: arXiv:2007.07021 [stat.ME]
  (or arXiv:2007.07021v5 [stat.ME] for this version)

Submission history

From: Ray Bai [view email]
[v1] Tue, 14 Jul 2020 13:29:13 GMT (54kb,D)
[v2] Sun, 26 Jul 2020 19:57:31 GMT (55kb,D)
[v3] Wed, 29 Jul 2020 06:13:38 GMT (55kb,D)
[v4] Mon, 21 Sep 2020 16:57:47 GMT (57kb,D)
[v5] Sat, 5 Jun 2021 15:22:40 GMT (101kb,D)

Link back to: arXiv, form interface, contact.