Timezone: »

Private Reinforcement Learning with PAC and Regret Guarantees
Giuseppe Vietri · Borja de Balle Pigem · Akshay Krishnamurthy · Steven Wu

Tue Jul 14 07:00 AM -- 07:45 AM & Tue Jul 14 08:00 PM -- 08:45 PM (PDT) @ Virtual

Motivated by high-stakes decision-making domains like personalized medicine where user information is inherently sensitive, we design privacy preserving exploration policies for episodic reinforcement learning (RL). We first provide a meaningful privacy formulation using the notion of joint differential privacy (JDP)--a strong variant of differential privacy for settings where each user receives their own sets of output (e.g., policy recommendations). We then develop a private optimism-based learning algorithm that simultaneously achieves strong PAC and regret bounds, and enjoys a JDP guarantee. Our algorithm only pays for a moderate privacy cost on exploration: in comparison to the non-private bounds, the privacy parameter only appears in lower-order terms. Finally, we present lower bounds on sample complexity and regret for reinforcement learning subject to JDP.

Author Information

Giuseppe Vietri (University of Minnesota)
Borja de Balle Pigem (Amazon Research)
Akshay Krishnamurthy (Microsoft Research)
Steven Wu (University of Minnesota)

More from the Same Authors