Current browse context:
cs.CR
Change to browse by:
References & Citations
Computer Science > Machine Learning
Title: On Private and Robust Bandits
(Submitted on 6 Feb 2023 (v1), last revised 4 Mar 2023 (this version, v2))
Abstract: We study private and robust multi-armed bandits (MABs), where the agent receives Huber's contaminated heavy-tailed rewards and meanwhile needs to ensure differential privacy. We first present its minimax lower bound, characterizing the information-theoretic limit of regret with respect to privacy budget, contamination level and heavy-tailedness. Then, we propose a meta-algorithm that builds on a private and robust mean estimation sub-routine \texttt{PRM} that essentially relies on reward truncation and the Laplace mechanism only. For two different heavy-tailed settings, we give specific schemes of \texttt{PRM}, which enable us to achieve nearly-optimal regret. As by-products of our main results, we also give the first minimax lower bound for private heavy-tailed MABs (i.e., without contamination). Moreover, our two proposed truncation-based \texttt{PRM} achieve the optimal trade-off between estimation accuracy, privacy and robustness. Finally, we support our theoretical results with experimental studies.
Submission history
From: Yulian Wu [view email][v1] Mon, 6 Feb 2023 01:55:06 GMT (667kb,D)
[v2] Sat, 4 Mar 2023 11:27:16 GMT (668kb,D)
Link back to: arXiv, form interface, contact.