Optimal Multi-Fidelity Best-Arm Identification

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Poiani, Riccardo, Degenne, Rémy, Kaufmann, Emilie, Metelli, Alberto Maria, Restelli, Marcello
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866910968070537216
author Poiani, Riccardo
Degenne, Rémy
Kaufmann, Emilie
Metelli, Alberto Maria
Restelli, Marcello
author_facet Poiani, Riccardo
Degenne, Rémy
Kaufmann, Emilie
Metelli, Alberto Maria
Restelli, Marcello
contents In bandit best-arm identification, an algorithm is tasked with finding the arm with highest mean reward with a specified accuracy as fast as possible. We study multi-fidelity best-arm identification, in which the algorithm can choose to sample an arm at a lower fidelity (less accurate mean estimate) for a lower cost. Several methods have been proposed for tackling this problem, but their optimality remain elusive, notably due to loose lower bounds on the total cost needed to identify the best arm. Our first contribution is a tight, instance-dependent lower bound on the cost complexity. The study of the optimization problem featured in the lower bound provides new insights to devise computationally efficient algorithms, and leads us to propose a gradient-based approach with asymptotically optimal cost complexity. We demonstrate the benefits of the new algorithm compared to existing methods in experiments. Our theoretical and empirical findings also shed light on an intriguing concept of optimal fidelity for each arm.
format Preprint
id arxiv_https___arxiv_org_abs_2406_03033
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Optimal Multi-Fidelity Best-Arm Identification
Poiani, Riccardo
Degenne, Rémy
Kaufmann, Emilie
Metelli, Alberto Maria
Restelli, Marcello
Machine Learning
In bandit best-arm identification, an algorithm is tasked with finding the arm with highest mean reward with a specified accuracy as fast as possible. We study multi-fidelity best-arm identification, in which the algorithm can choose to sample an arm at a lower fidelity (less accurate mean estimate) for a lower cost. Several methods have been proposed for tackling this problem, but their optimality remain elusive, notably due to loose lower bounds on the total cost needed to identify the best arm. Our first contribution is a tight, instance-dependent lower bound on the cost complexity. The study of the optimization problem featured in the lower bound provides new insights to devise computationally efficient algorithms, and leads us to propose a gradient-based approach with asymptotically optimal cost complexity. We demonstrate the benefits of the new algorithm compared to existing methods in experiments. Our theoretical and empirical findings also shed light on an intriguing concept of optimal fidelity for each arm.
title Optimal Multi-Fidelity Best-Arm Identification
topic Machine Learning
url https://arxiv.org/abs/2406.03033