We introduce the Bayesian Nested Group Lasso, a hierarchical model extending the Group Lasso to nested structures. By formulating the penalty as a scale mixture of Gaussians, we derive an efficient MCMC sampling scheme. A simulation study shows that our method reduces posterior variability for irrelevant variables, outperforming the Bayesian Lasso in structured sparsity settings. We discuss potential extensions, including spike-and-slab priors, for exact variable selection.

Stefanucci, M., Alaimo Di Loro, P., Barone, R., Bayesian Nested Group Lasso, in Statistics for Innovation IV, (Genova, 16-18 June 2025), Springer, Genova 2025: 369-374. [10.1007/978-3-031-96033-8_60] [https://hdl.handle.net/10807/323989]

Bayesian Nested Group Lasso

Barone, Rosario
2025

Abstract

We introduce the Bayesian Nested Group Lasso, a hierarchical model extending the Group Lasso to nested structures. By formulating the penalty as a scale mixture of Gaussians, we derive an efficient MCMC sampling scheme. A simulation study shows that our method reduces posterior variability for irrelevant variables, outperforming the Bayesian Lasso in structured sparsity settings. We discuss potential extensions, including spike-and-slab priors, for exact variable selection.
2025
Inglese
Statistics for Innovation IV
SIS 2025 – Statistics for Innovation
Genova
16-giu-2025
18-giu-2025
978-3-031-96035-2
Springer
Stefanucci, M., Alaimo Di Loro, P., Barone, R., Bayesian Nested Group Lasso, in Statistics for Innovation IV, (Genova, 16-18 June 2025), Springer, Genova 2025: 369-374. [10.1007/978-3-031-96033-8_60] [https://hdl.handle.net/10807/323989]
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10807/323989
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? 0
social impact