TrajDeleter: Enabling Trajectory Forgetting in Offline Reinforcement Learning Agents
Chen Gong
Network and Distributed System Security (NDSS) Symposium 2025 · Day 3 · Machine Unlearning
This talk introduces **TrajDeleter**, a novel framework designed to enable efficient and stable trajectory forgetting in **offline reinforcement learning (RL)** agents. Presented by Chen Gong from Kung Fu University Virginia, the research addresses a critical and increasingly relevant challenge in the era of data privacy and regulatory compliance: the "right to be forgotten" in machine learning models, specifically within the context of RL agents trained on fixed datasets of past interactions. The inability to selectively remove the influence of specific data points (trajectories) from a trained RL agent without costly retraining poses significant hurdles for applications ranging from autonomous medical treatment plans that handle sensitive patient information to robotics that might inadvertently store private environmental layouts, and adherence to regulations like **GDPR**.
AI review
Legitimate academic ML research solving a real problem — selective trajectory unlearning in offline RL with a two-phase approach and MIA-based auditing. Solid contribution to a niche but growing area, with honest acknowledgment of limitations. Not a security talk in any traditional sense, but it belongs at a venue like NDSS where the privacy/ML intersection lives.