OPTAMI: Global Superlinear Convergence of High-order Methods

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Kamzolov, Dmitry, Pasechnyuk, Dmitry, Agafonov, Artem, Gasnikov, Alexander, Takáč, Martin
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866917801172664320
author Kamzolov, Dmitry
Pasechnyuk, Dmitry
Agafonov, Artem
Gasnikov, Alexander
Takáč, Martin
author_facet Kamzolov, Dmitry
Pasechnyuk, Dmitry
Agafonov, Artem
Gasnikov, Alexander
Takáč, Martin
contents Second-order methods for convex optimization outperform first-order methods in terms of theoretical iteration convergence, achieving rates up to $O(k^{-5})$ for highly-smooth functions. However, their practical performance and applications are limited due to their multi-level structure and implementation complexity. In this paper, we present new results on high-order optimization methods, supported by their practical performance. First, we show that the basic high-order methods, such as the Cubic Regularized Newton Method, exhibit global superlinear convergence for $μ$-strongly star-convex functions, a class that includes $μ$-strongly convex functions and some non-convex functions. Theoretical convergence results are both inspired and supported by the practical performance of these methods. Secondly, we propose a practical version of the Nesterov Accelerated Tensor method, called NATA. It significantly outperforms the classical variant and other high-order acceleration techniques in practice. The convergence of NATA is also supported by theoretical results. Finally, we introduce an open-source computational library for high-order methods, called OPTAMI. This library includes various methods, acceleration techniques, and subproblem solvers, all implemented as PyTorch optimizers, thereby facilitating the practical application of high-order methods to a wide range of optimization problems. We hope this library will simplify research and practical comparison of methods beyond first-order.
format Preprint
id arxiv_https___arxiv_org_abs_2410_04083
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle OPTAMI: Global Superlinear Convergence of High-order Methods
Kamzolov, Dmitry
Pasechnyuk, Dmitry
Agafonov, Artem
Gasnikov, Alexander
Takáč, Martin
Optimization and Control
Second-order methods for convex optimization outperform first-order methods in terms of theoretical iteration convergence, achieving rates up to $O(k^{-5})$ for highly-smooth functions. However, their practical performance and applications are limited due to their multi-level structure and implementation complexity. In this paper, we present new results on high-order optimization methods, supported by their practical performance. First, we show that the basic high-order methods, such as the Cubic Regularized Newton Method, exhibit global superlinear convergence for $μ$-strongly star-convex functions, a class that includes $μ$-strongly convex functions and some non-convex functions. Theoretical convergence results are both inspired and supported by the practical performance of these methods. Secondly, we propose a practical version of the Nesterov Accelerated Tensor method, called NATA. It significantly outperforms the classical variant and other high-order acceleration techniques in practice. The convergence of NATA is also supported by theoretical results. Finally, we introduce an open-source computational library for high-order methods, called OPTAMI. This library includes various methods, acceleration techniques, and subproblem solvers, all implemented as PyTorch optimizers, thereby facilitating the practical application of high-order methods to a wide range of optimization problems. We hope this library will simplify research and practical comparison of methods beyond first-order.
title OPTAMI: Global Superlinear Convergence of High-order Methods
topic Optimization and Control
url https://arxiv.org/abs/2410.04083