Batched Stochastic Bandit for Nondegenerate Functions

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Yu, Shu, Yunlu, Wang, Tianyu
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912314850017280
author Liu, Yu
Shu, Yunlu
Wang, Tianyu
author_facet Liu, Yu
Shu, Yunlu
Wang, Tianyu
contents This paper studies batched bandit learning problems for nondegenerate functions. We introduce an algorithm that solves the batched bandit problem for nondegenerate functions near-optimally. More specifically, we introduce an algorithm, called Geometric Narrowing (GN), whose regret bound is of order $\widetilde{\mathcal{O}} ( A_{+}^d \sqrt{T} )$. In addition, GN only needs $\mathcal{O} (\log \log T)$ batches to achieve this regret. We also provide lower bound analysis for this problem. More specifically, we prove that over some (compact) doubling metric space of doubling dimension $d$: 1. For any policy $π$, there exists a problem instance on which $π$ admits a regret of order $Ω ( A_-^d \sqrt{T})$; 2. No policy can achieve a regret of order $ A_-^d \sqrt{T} $ over all problem instances, using less than $ Ω( \log \log T ) $ rounds of communications. Our lower bound analysis shows that the GN algorithm achieves near optimal regret with minimal number of batches.
format Preprint
id arxiv_https___arxiv_org_abs_2405_05733
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Batched Stochastic Bandit for Nondegenerate Functions
Liu, Yu
Shu, Yunlu
Wang, Tianyu
Machine Learning
This paper studies batched bandit learning problems for nondegenerate functions. We introduce an algorithm that solves the batched bandit problem for nondegenerate functions near-optimally. More specifically, we introduce an algorithm, called Geometric Narrowing (GN), whose regret bound is of order $\widetilde{\mathcal{O}} ( A_{+}^d \sqrt{T} )$. In addition, GN only needs $\mathcal{O} (\log \log T)$ batches to achieve this regret. We also provide lower bound analysis for this problem. More specifically, we prove that over some (compact) doubling metric space of doubling dimension $d$: 1. For any policy $π$, there exists a problem instance on which $π$ admits a regret of order $Ω ( A_-^d \sqrt{T})$; 2. No policy can achieve a regret of order $ A_-^d \sqrt{T} $ over all problem instances, using less than $ Ω( \log \log T ) $ rounds of communications. Our lower bound analysis shows that the GN algorithm achieves near optimal regret with minimal number of batches.
title Batched Stochastic Bandit for Nondegenerate Functions
topic Machine Learning
url https://arxiv.org/abs/2405.05733