Sparse parametric models are of great interest in statistical learning and are often analyzed by means of regularized estimators. Pathwise methods allow us to efficiently compute the full solution path for penalized estimators, for any possible value of the penalization parameter lambda\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\lambda $$\end{document}. In this paper we deal with the pathwise optimization for bridge-type problems; i.e., we are interested in the minimization of a loss function, such as negative log-likelihood or residual sum of squares, plus the sum of & ell;q\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\ell <^>q$$\end{document} norms with q is an element of(0,1]\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$q\in (0,1]$$\end{document} involving adaptive coefficients. For some loss functions this regularization achieves asymptotically the oracle properties (such as selection consistency). Nevertheless, since the objective function involves non-convex and non-differentiable terms, the minimization problem is computationally challenging. The aim of this paper is to apply some general algorithms, arising from non-convex optimization theory, to compute efficiently the path solutions for the adaptive bridge estimator with multiple penalties. In particular, we take into account two different approaches: accelerated proximal gradient descent and blockwise alternating optimization. A pathwise scheme not only enables rapid, grid-based validation of the tuning parameter but also helps avoid spurious local minima in the underlying nonconvex optimization. The convergence and the path consistency of these algorithms are discussed. In order to assess our methods, we apply these algorithms to the penalized estimation of diffusion processes observed at discrete times. The latter represents a recent research topic in the field of statistics for time-dependent data.

Pathwise optimization for bridge-type estimators and its applications / De Gregorio, A., Iafrate, F.. - In: STATISTICS AND COMPUTING. - ISSN 0960-3174. - 36:3(2026), pp. 1-20. [10.1007/s11222-026-10856-2]

Pathwise optimization for bridge-type estimators and its applications

De Gregorio A.
;
2026

Abstract

Sparse parametric models are of great interest in statistical learning and are often analyzed by means of regularized estimators. Pathwise methods allow us to efficiently compute the full solution path for penalized estimators, for any possible value of the penalization parameter lambda\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\lambda $$\end{document}. In this paper we deal with the pathwise optimization for bridge-type problems; i.e., we are interested in the minimization of a loss function, such as negative log-likelihood or residual sum of squares, plus the sum of & ell;q\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\ell <^>q$$\end{document} norms with q is an element of(0,1]\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$q\in (0,1]$$\end{document} involving adaptive coefficients. For some loss functions this regularization achieves asymptotically the oracle properties (such as selection consistency). Nevertheless, since the objective function involves non-convex and non-differentiable terms, the minimization problem is computationally challenging. The aim of this paper is to apply some general algorithms, arising from non-convex optimization theory, to compute efficiently the path solutions for the adaptive bridge estimator with multiple penalties. In particular, we take into account two different approaches: accelerated proximal gradient descent and blockwise alternating optimization. A pathwise scheme not only enables rapid, grid-based validation of the tuning parameter but also helps avoid spurious local minima in the underlying nonconvex optimization. The convergence and the path consistency of these algorithms are discussed. In order to assess our methods, we apply these algorithms to the penalized estimation of diffusion processes observed at discrete times. The latter represents a recent research topic in the field of statistics for time-dependent data.
2026
adaptive thresholding operators; gradient descent; non-convex optimization; path consistency; proximal maps; regularized estimators; stochastic differential equations
01 Pubblicazione su rivista::01a Articolo in rivista
Pathwise optimization for bridge-type estimators and its applications / De Gregorio, A., Iafrate, F.. - In: STATISTICS AND COMPUTING. - ISSN 0960-3174. - 36:3(2026), pp. 1-20. [10.1007/s11222-026-10856-2]
File allegati a questo prodotto
File Dimensione Formato  
De Gregorio_Pathwise-optimization_2026.pdf

accesso aperto

Tipologia: Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza: Tutti i diritti riservati (All rights reserved)
Dimensione 1.32 MB
Formato Adobe PDF
1.32 MB Adobe PDF

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/11573/1775882
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? 0
social impact