CLARSTA: A random subspace trust-region algorithm for convex-constrained derivative-free optimization

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Chen, Yiwen, Hare, Warren, Wiebe, Amy
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866917487413559296
author Chen, Yiwen
Hare, Warren
Wiebe, Amy
author_facet Chen, Yiwen
Hare, Warren
Wiebe, Amy
contents This paper proposes a random subspace trust-region algorithm for general convex-constrained derivative-free optimization (DFO) problems. Similar to previous random subspace DFO methods, the convergence of our algorithm requires a certain accuracy of models and a certain quality of subspaces. For model accuracy, we define a new class of models that is only required to provide reasonable accuracy on the projection of the constraint set onto the subspace. We provide a new geometry measure to make these models easy to analyze, construct, and manage. For subspace quality, we use the concentration of measure on the Grassmann manifold to provide a method to sample subspaces that preserve the first-order criticality measure by a certain fraction with a certain probability lower bound. Based on all these new theoretical results, we present an almost-sure global convergence and a worst-case complexity analysis of our algorithm. Numerical experiments on problems with dimensions up to 10000 demonstrate the reliable performance of our algorithm in high dimensions.
format Preprint
id arxiv_https___arxiv_org_abs_2506_20335
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CLARSTA: A random subspace trust-region algorithm for convex-constrained derivative-free optimization
Chen, Yiwen
Hare, Warren
Wiebe, Amy
Optimization and Control
90C56, 65K05, 90C06
This paper proposes a random subspace trust-region algorithm for general convex-constrained derivative-free optimization (DFO) problems. Similar to previous random subspace DFO methods, the convergence of our algorithm requires a certain accuracy of models and a certain quality of subspaces. For model accuracy, we define a new class of models that is only required to provide reasonable accuracy on the projection of the constraint set onto the subspace. We provide a new geometry measure to make these models easy to analyze, construct, and manage. For subspace quality, we use the concentration of measure on the Grassmann manifold to provide a method to sample subspaces that preserve the first-order criticality measure by a certain fraction with a certain probability lower bound. Based on all these new theoretical results, we present an almost-sure global convergence and a worst-case complexity analysis of our algorithm. Numerical experiments on problems with dimensions up to 10000 demonstrate the reliable performance of our algorithm in high dimensions.
title CLARSTA: A random subspace trust-region algorithm for convex-constrained derivative-free optimization
topic Optimization and Control
90C56, 65K05, 90C06
url https://arxiv.org/abs/2506.20335