Optimal Thresholding Linear Bandit

arXiv:2402.09467v1 Announce Type: cross
Abstract: We study a novel pure exploration problem: the $\epsilon$-Thresholding Bandit Problem (TBP) with fixed confidence in stochastic linear bandits. We prove a lower bound for the sample complexity and extend an algorithm designed for Best Arm Identification in the linear case to TBP that is asymptotically optimal.

abstract algorithm arm arxiv case complexity confidence cs.lg exploration identification linear novel prove sample stat.ml stochastic study thresholding type