Skip to Main Content (Press Enter)

Logo UNIMI
  • ×
  • Home
  • Persone
  • Attività
  • Ambiti
  • Strutture
  • Pubblicazioni
  • Terza Missione

Expertise & Skills
Logo UNIMI

|

Expertise & Skills

unimi.it
  • ×
  • Home
  • Persone
  • Attività
  • Ambiti
  • Strutture
  • Pubblicazioni
  • Terza Missione
  1. Pubblicazioni

Sparsity-Agnostic Linear Bandits with Adaptive Adversaries

Contributo in Atti di convegno
Data di Pubblicazione:
2024
Citazione:
Sparsity-Agnostic Linear Bandits with Adaptive Adversaries / T. Jin, K. Jang, N. Cesa Bianchi (ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS). - In: Advances in Neural Information Processing Systems / [a cura di] A. Globerson and L. Mackey and D. Belgrave and A. Fan and U. Paquet and J. Tomczak and C. Zhang. - [s.l] : Curran Associates, Inc., 2024. - ISBN 9798331314385. - pp. 42015-42047 (( Intervento presentato al 38. convegno Annual Conference on Neural Information Processing Systems tenutosi a Vancouver nel 2024.
Abstract:
We study stochastic linear bandits where, in each round, the learner receives a set of actions (i.e., feature vectors), from which it chooses an element and obtains a stochastic reward. The expected reward is a fixed but unknown linear function of the chosen action. We study sparse regret bounds, that depend on the number S of non-zero coefficients in the linear reward function. Previous works focused on the case where S is known, or the action sets satisfy additional assumptions. In this work, we obtain the first sparse regret bounds that hold when S is unknown and the action sets are adversarially generated. Our techniques combine online to confidence set conversions with a novel randomized model selection approach over a hierarchy of nested confidence sets. When S is known, our analysis recovers state-of-the-art bounds for adversarial action sets. We also show that a variant of our approach, using Exp3 to dynamically select the confidence sets, can be used to improve the empirical performance of stochastic linear bandits while enjoying a regret bound with optimal dependence on the time horizon.
Tipologia IRIS:
03 - Contributo in volume
Elenco autori:
T. Jin, K. Jang, N. Cesa Bianchi
Autori di Ateneo:
CESA BIANCHI NICOLO' ANTONIO ( autore )
Link alla scheda completa:
https://air.unimi.it/handle/2434/1157683
Link al Full Text:
https://air.unimi.it/retrieve/handle/2434/1157683/2776279/NeurIPS-2024-sparsity-agnostic-linear-bandits-with-adaptive-adversaries-Paper-Conference.pdf
Titolo del libro:
Advances in Neural Information Processing Systems
Progetto:
Learning in Markets and Society
  • Aree Di Ricerca

Aree Di Ricerca

Settori


Settore INFO-01/A - Informatica
  • Informazioni
  • Assistenza
  • Accessibilità
  • Privacy
  • Utilizzo dei cookie
  • Note legali

Realizzato con VIVO | Progettato da Cineca | 25.11.5.0