Skip the Hessian, Keep the Rates: Globalized Semismooth Newton with Lazy Hessian Updates

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Alphonse, Amal, Dvurechensky, Pavel, Sirotenko, Clemens
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908820608909312
author Alphonse, Amal
Dvurechensky, Pavel
Sirotenko, Clemens
author_facet Alphonse, Amal
Dvurechensky, Pavel
Sirotenko, Clemens
contents Second-order methods are provably faster than first-order methods, and their efficient implementations for large-scale optimization problems have attracted significant attention. Yet, optimization problems in ML often have nonsmooth derivatives, which makes the existing convergence rate theory of second-order methods inapplicable. In this paper, we propose a new semismooth Newton method (SSN) that enjoys both global convergence rates and asymptotic superlinear convergence without requiring second-order differentiability. Crucially, our method does not require (generalized) Hessians to be evaluated at each iteration but only periodically, and it reuses stale Hessians otherwise (i.e., it performs lazy Hessian updates), saving compute cost and often leading to significant speedups in time, whilst still maintaining strong global and local convergence rate guarantees. We develop our theory in an infinite-dimensional setting and illustrate it with numerical experiments on matrix factorization and neural networks with Lipschitz constraints.
format Preprint
id arxiv_https___arxiv_org_abs_2602_08069
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Skip the Hessian, Keep the Rates: Globalized Semismooth Newton with Lazy Hessian Updates
Alphonse, Amal
Dvurechensky, Pavel
Sirotenko, Clemens
Optimization and Control
Second-order methods are provably faster than first-order methods, and their efficient implementations for large-scale optimization problems have attracted significant attention. Yet, optimization problems in ML often have nonsmooth derivatives, which makes the existing convergence rate theory of second-order methods inapplicable. In this paper, we propose a new semismooth Newton method (SSN) that enjoys both global convergence rates and asymptotic superlinear convergence without requiring second-order differentiability. Crucially, our method does not require (generalized) Hessians to be evaluated at each iteration but only periodically, and it reuses stale Hessians otherwise (i.e., it performs lazy Hessian updates), saving compute cost and often leading to significant speedups in time, whilst still maintaining strong global and local convergence rate guarantees. We develop our theory in an infinite-dimensional setting and illustrate it with numerical experiments on matrix factorization and neural networks with Lipschitz constraints.
title Skip the Hessian, Keep the Rates: Globalized Semismooth Newton with Lazy Hessian Updates
topic Optimization and Control
url https://arxiv.org/abs/2602.08069