References & Citations
Statistics > Methodology
Title: Bayesian Generalized Additive Model Selection Including a Fast Variational Option
(Submitted on 2 Jan 2022 (v1), last revised 28 Sep 2023 (this version, v6))
Abstract: We use Bayesian model selection paradigms, such as group least absolute shrinkage and selection operator priors, to facilitate generalized additive model selection. Our approach allows for the effects of continuous predictors to be categorized as either zero, linear or non-linear. Employment of carefully tailored auxiliary variables results in Gibbsian Markov chain Monte Carlo schemes for practical implementation of the approach. In addition, mean field variational algorithms with closed form updates are obtained. Whilst not as accurate, this fast variational option enhances scalability to very large data sets. A package in the R language aids use in practice.
Submission history
From: Matt Wand [view email][v1] Sun, 2 Jan 2022 20:16:19 GMT (110kb,D)
[v2] Tue, 1 Feb 2022 06:21:31 GMT (110kb,D)
[v3] Wed, 18 Jan 2023 22:19:48 GMT (110kb,D)
[v4] Wed, 5 Apr 2023 03:56:41 GMT (176kb,D)
[v5] Thu, 31 Aug 2023 20:49:40 GMT (205kb,D)
[v6] Thu, 28 Sep 2023 05:43:49 GMT (205kb,D)
Link back to: arXiv, form interface, contact.