Minimax Optimal Fixed-Budget Best Arm Identification in Linear BanditsDownload PDF

Published: 31 Oct 2022, 18:00, Last Modified: 09 Jan 2023, 10:04NeurIPS 2022 AcceptReaders: Everyone
Keywords: best arm identification, linear bandits
TL;DR: A minimax optimal algorithm for best arm identification in linear bandits in the fixed-budget setting.
Abstract: We study the problem of best arm identification in linear bandits in the fixed-budget setting. By leveraging properties of the G-optimal design and incorporating it into the arm allocation rule, we design a parameter-free algorithm, Optimal Design-based Linear Best Arm Identification (OD-LinBAI). We provide a theoretical analysis of the failure probability of OD-LinBAI. Instead of all the optimality gaps, the performance of OD-LinBAI depends only on the gaps of the top $d$ arms, where $d$ is the effective dimension of the linear bandit instance. Complementarily, we present a minimax lower bound for this problem. The upper and lower bounds show that OD-LinBAI is minimax optimal up to constant multiplicative factors in the exponent, which is a significant theoretical improvement over existing methods (e.g., BayesGap, Peace, LinearExploration and GSE), and settles the question of ascertaining the difficulty of learning the best arm in the fixed-budget setting. Finally, numerical experiments demonstrate considerable empirical improvements over existing algorithms on a variety of real and synthetic datasets.
Supplementary Material: zip
11 Replies