Saved in:
Bibliographic Details
Main Authors: Entezari, Alireza, Banerjee, Arunava
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.09469
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • Current performance bounds for randomized iterative methods are often considered tight under per-iteration analyses, yet they are notoriously loose in practice. We derive asymptotic performance bounds that narrow this theory-practice gap, leveraging a new technique for bounding the spectral radii of operators arising in randomized iterations and a connection we establish to Perron-Frobenius theory for noncommutative algebras. The asymptotic analysis also uncovers and quantifies the previously unexplained role of relaxation in improving performance, thereby resolving an open problem posed by Strohmer and Vershynin in 2007.