Affine-Invariant Global Non-Asymptotic Convergence Analysis of BFGS under Self-Concordance

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Jin, Qiujiang, Mokhtari, Aryan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915575610998784
author Jin, Qiujiang
Mokhtari, Aryan
author_facet Jin, Qiujiang
Mokhtari, Aryan
contents In this paper, we establish global non-asymptotic convergence guarantees for the BFGS quasi-Newton method without requiring strong convexity or the Lipschitz continuity of the gradient or Hessian. Instead, we consider the setting where the objective function is strictly convex and strongly self-concordant. For an arbitrary initial point and any arbitrary positive-definite initial Hessian approximation, we prove global linear and superlinear convergence guarantees for BFGS when the step size is determined using a line search scheme satisfying the weak Wolfe conditions. Moreover, all our global guarantees are affine-invariant, with the convergence rates depending solely on the initial error and the strongly self-concordant constant. Our results extend the global non-asymptotic convergence theory of BFGS beyond traditional assumptions and, for the first time, establish affine-invariant convergence guarantees aligning with the inherent affine invariance of the BFGS method.
format Preprint
id arxiv_https___arxiv_org_abs_2507_00361
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Affine-Invariant Global Non-Asymptotic Convergence Analysis of BFGS under Self-Concordance
Jin, Qiujiang
Mokhtari, Aryan
Optimization and Control
Numerical Analysis
In this paper, we establish global non-asymptotic convergence guarantees for the BFGS quasi-Newton method without requiring strong convexity or the Lipschitz continuity of the gradient or Hessian. Instead, we consider the setting where the objective function is strictly convex and strongly self-concordant. For an arbitrary initial point and any arbitrary positive-definite initial Hessian approximation, we prove global linear and superlinear convergence guarantees for BFGS when the step size is determined using a line search scheme satisfying the weak Wolfe conditions. Moreover, all our global guarantees are affine-invariant, with the convergence rates depending solely on the initial error and the strongly self-concordant constant. Our results extend the global non-asymptotic convergence theory of BFGS beyond traditional assumptions and, for the first time, establish affine-invariant convergence guarantees aligning with the inherent affine invariance of the BFGS method.
title Affine-Invariant Global Non-Asymptotic Convergence Analysis of BFGS under Self-Concordance
topic Optimization and Control
Numerical Analysis
url https://arxiv.org/abs/2507.00361