Randomized coordinate gradient descent almost surely escapes strict saddle points

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Ziang, Li, Yingzhou, Li, Zihao
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913983304302592
author Chen, Ziang
Li, Yingzhou
Li, Zihao
author_facet Chen, Ziang
Li, Yingzhou
Li, Zihao
contents We analyze the behavior of randomized coordinate gradient descent for nonconvex optimization, proving that under standard assumptions, the iterates almost surely escape strict saddle points. By formulating the method as a nonlinear random dynamical system and characterizing neighborhoods of critical points, we establish this result through the center-stable manifold theorem.
format Preprint
id arxiv_https___arxiv_org_abs_2508_07535
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Randomized coordinate gradient descent almost surely escapes strict saddle points
Chen, Ziang
Li, Yingzhou
Li, Zihao
Optimization and Control
Numerical Analysis
Dynamical Systems
Probability
We analyze the behavior of randomized coordinate gradient descent for nonconvex optimization, proving that under standard assumptions, the iterates almost surely escape strict saddle points. By formulating the method as a nonlinear random dynamical system and characterizing neighborhoods of critical points, we establish this result through the center-stable manifold theorem.
title Randomized coordinate gradient descent almost surely escapes strict saddle points
topic Optimization and Control
Numerical Analysis
Dynamical Systems
Probability
url https://arxiv.org/abs/2508.07535