EVaR-Optimal Arm Identification in Bandits
Fuente:
arXiv
Salvato in:
| Autori principali: | , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916991300796416 |
|---|---|
| author | Ahmadipour, Mehrasa Garivier, Aurélien |
| author_facet | Ahmadipour, Mehrasa Garivier, Aurélien |
| contents | We study the fixed-confidence best arm identification (BAI) problem within the multi-armed bandit (MAB) framework under the Entropic Value-at-Risk (EVaR) criterion. Our analysis considers a nonparametric setting, allowing for general reward distributions bounded in [0,1]. This formulation addresses the critical need for risk-averse decision-making in high-stakes environments, such as finance, moving beyond simple expected value optimization. We propose a $δ$-correct, Track-and-Stop based algorithm and derive a corresponding lower bound on the expected sample complexity, which we prove is asymptotically matched. The implementation of our algorithm and the characterization of the lower bound both require solving a complex convex optimization problem and a related, simpler non-convex one. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_04728 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | EVaR-Optimal Arm Identification in Bandits Ahmadipour, Mehrasa Garivier, Aurélien Machine Learning We study the fixed-confidence best arm identification (BAI) problem within the multi-armed bandit (MAB) framework under the Entropic Value-at-Risk (EVaR) criterion. Our analysis considers a nonparametric setting, allowing for general reward distributions bounded in [0,1]. This formulation addresses the critical need for risk-averse decision-making in high-stakes environments, such as finance, moving beyond simple expected value optimization. We propose a $δ$-correct, Track-and-Stop based algorithm and derive a corresponding lower bound on the expected sample complexity, which we prove is asymptotically matched. The implementation of our algorithm and the characterization of the lower bound both require solving a complex convex optimization problem and a related, simpler non-convex one. |
| title | EVaR-Optimal Arm Identification in Bandits |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2510.04728 |