A Study of Condition Numbers for First-Order Optimization

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Guille-Escuret, Charles, Goujaud, Baptiste, Girotti, Manuela, Mitliagkas, Ioannis
Formato: Preprint
Publicado: 2020
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866929362562973696
author Guille-Escuret, Charles
Goujaud, Baptiste
Girotti, Manuela
Mitliagkas, Ioannis
author_facet Guille-Escuret, Charles
Goujaud, Baptiste
Girotti, Manuela
Mitliagkas, Ioannis
contents The study of first-order optimization algorithms (FOA) typically starts with assumptions on the objective functions, most commonly smoothness and strong convexity. These metrics are used to tune the hyperparameters of FOA. We introduce a class of perturbations quantified via a new norm, called *-norm. We show that adding a small perturbation to the objective function has an equivalently small impact on the behavior of any FOA, which suggests that it should have a minor impact on the tuning of the algorithm. However, we show that smoothness and strong convexity can be heavily impacted by arbitrarily small perturbations, leading to excessively conservative tunings and convergence issues. In view of these observations, we propose a notion of continuity of the metrics, which is essential for a robust tuning strategy. Since smoothness and strong convexity are not continuous, we propose a comprehensive study of existing alternative metrics which we prove to be continuous. We describe their mutual relations and provide their guaranteed convergence rates for the Gradient Descent algorithm accordingly tuned. Finally we discuss how our work impacts the theoretical understanding of FOA and their performances.
format Preprint
id arxiv_https___arxiv_org_abs_2012_05782
institution arXiv
publishDate 2020
record_format arxiv
spellingShingle A Study of Condition Numbers for First-Order Optimization
Guille-Escuret, Charles
Goujaud, Baptiste
Girotti, Manuela
Mitliagkas, Ioannis
Machine Learning
Optimization and Control
The study of first-order optimization algorithms (FOA) typically starts with assumptions on the objective functions, most commonly smoothness and strong convexity. These metrics are used to tune the hyperparameters of FOA. We introduce a class of perturbations quantified via a new norm, called *-norm. We show that adding a small perturbation to the objective function has an equivalently small impact on the behavior of any FOA, which suggests that it should have a minor impact on the tuning of the algorithm. However, we show that smoothness and strong convexity can be heavily impacted by arbitrarily small perturbations, leading to excessively conservative tunings and convergence issues. In view of these observations, we propose a notion of continuity of the metrics, which is essential for a robust tuning strategy. Since smoothness and strong convexity are not continuous, we propose a comprehensive study of existing alternative metrics which we prove to be continuous. We describe their mutual relations and provide their guaranteed convergence rates for the Gradient Descent algorithm accordingly tuned. Finally we discuss how our work impacts the theoretical understanding of FOA and their performances.
title A Study of Condition Numbers for First-Order Optimization
topic Machine Learning
Optimization and Control
url https://arxiv.org/abs/2012.05782