BriefGPT.xyz
Sep, 2022
UCB基于最佳臂识别策略的对抗攻击的样本复杂度
Sample Complexity of an Adversarial Attack on UCB-based Best-arm Identification Policy
HTML
PDF
Varsha Pendyala
TL;DR
研究多臂赌博机中UCB类型最佳臂识别策略的对手扰动攻击,探讨其对模型选择目标臂的影响,证明了在总臂数和伪中心极限定理参数已知的情况下,可以在T轮内找到目标臂作为最佳臂。
Abstract
In this work I study the problem of
adversarial perturbations
to rewards, in a
multi-armed bandit
(MAB) setting. Specifically, I focus on an adversarial attack to a UCB type
→