Salvato in:
Dettagli Bibliografici
Autori principali: Yoon, TaeHo, Ryu, Ernest K., Grimmer, Benjamin
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2511.14915
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917090331459584
author Yoon, TaeHo
Ryu, Ernest K.
Grimmer, Benjamin
author_facet Yoon, TaeHo
Ryu, Ernest K.
Grimmer, Benjamin
contents For nonexpansive fixed-point problems, Halpern's method with optimal parameters, its so-called H-dual algorithm, and in fact, an infinite family of algorithms containing them, all exhibit the exactly minimax optimal convergence rates. In this work, we provide a characterization of the complete, exhaustive family of distinct algorithms using predetermined step-sizes, represented as lower triangular H-matrices, which attain the same optimal convergence rate. The characterization is based on polynomials in the entries of the H-matrix that we call H-invariants, whose values stay constant over all optimal H-matrices, together with H-certificates, of which nonnegativity precisely specifies the region of optimality within the common level set of H-invariants. The H-invariance theory we present offers a novel view of optimal acceleration in first-order optimization as a mathematical study of carefully selected invariants, certificates, and structures induced by them.
format Preprint
id arxiv_https___arxiv_org_abs_2511_14915
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle H-invariance theory: A complete characterization of minimax optimal fixed-point algorithms
Yoon, TaeHo
Ryu, Ernest K.
Grimmer, Benjamin
Optimization and Control
For nonexpansive fixed-point problems, Halpern's method with optimal parameters, its so-called H-dual algorithm, and in fact, an infinite family of algorithms containing them, all exhibit the exactly minimax optimal convergence rates. In this work, we provide a characterization of the complete, exhaustive family of distinct algorithms using predetermined step-sizes, represented as lower triangular H-matrices, which attain the same optimal convergence rate. The characterization is based on polynomials in the entries of the H-matrix that we call H-invariants, whose values stay constant over all optimal H-matrices, together with H-certificates, of which nonnegativity precisely specifies the region of optimality within the common level set of H-invariants. The H-invariance theory we present offers a novel view of optimal acceleration in first-order optimization as a mathematical study of carefully selected invariants, certificates, and structures induced by them.
title H-invariance theory: A complete characterization of minimax optimal fixed-point algorithms
topic Optimization and Control
url https://arxiv.org/abs/2511.14915