Self-concordant smoothing in proximal quasi-Newton algorithms for large-scale convex composite optimization

Read full text Summarise with VIRAT
Authors: Adeyemi D. Adeoye, Alberto Bemporad
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

We introduce a notion of self-concordant smoothing for minimizing the sum of two convex functions, one of which is smooth and the other nonsmooth. The key highlight is a natural property of the resulting problem's structure that yields a variable metric selection method and a step length rule especially suited to proximal quasi-Newton algorithms. Also, we efficiently handle specific structures promoted by the nonsmooth term, such as ℓ1-regularization and group-lasso penalties. A convergence analysis for the class of proximal quasi-Newton methods covered by our framework is presented. In particular, we obtain guarantees, under standard assumptions, for two algorithms: Prox-N-SCORE (a proximal Newton method) and Prox-GGN-SCORE (a proximal generalized Gauss-Newton method). The latter uses a low-rank approximation of the Hessian inverse, reducing most of the cost of matrix inversion and making it effective for overparameterized machine learning models. Numerical experiments on synthetic and real data demonstrate the efficiency of both algorithms against state-of-the-art approaches. A Julia implementation is publicly available at https://github.com/adeyemiadeoye/SelfConcordantSmoothOptimization.jl.

Fields of Study & Subjects

Stochastic Gradient Optimization Techniques Advanced Optimization Algorithms Research Optimization and Variational Analysis

Publication details

Journal / source
Figshare
Publication date
30 March 2026
Language
EN