Local Differential Privacy for Regret Minimization in Reinforcement LearningDownload PDF

21 May 2021, 20:47 (modified: 26 Oct 2021, 11:54)NeurIPS 2021 PosterReaders: Everyone
Keywords: Regret, Optimism, Local Differential Privacy, Theory RL
TL;DR: We introduce the setting of Local Differential Privacy in RL, provide the first lower bound for this setting and present an algorithm with a regret matching the lower bound rate.
Abstract: Reinforcement learning algorithms are widely used in domains where it is desirable to provide a personalized service. In these domains it is common that user data contains sensitive information that needs to be protected from third parties. Motivated by this, we study privacy in the context of finite-horizon Markov Decision Processes (MDPs) by requiring information to be obfuscated on the user side. We formulate this notion of privacy for RL by leveraging the local differential privacy (LDP) framework. We establish a lower bound for regret minimization in finite-horizon MDPs with LDP guarantees which shows that guaranteeing privacy has a multiplicative effect on the regret. This result shows that while LDP is an appealing notion of privacy, it makes the learning problem significantly more complex. Finally, we present an optimistic algorithm that simultaneously satisfies $\varepsilon$-LDP requirements, and achieves $\sqrt{K}/\varepsilon$ regret in any finite-horizon MDP after $K$ episodes, matching the lower bound dependency on the number of episodes $K$.
Supplementary Material: pdf
Code Of Conduct: I certify that all co-authors of this work have read and commit to adhering to the NeurIPS Statement on Ethics, Fairness, Inclusivity, and Code of Conduct.
12 Replies