Policy Poisoning In Batch Reinforcement Learning And Control

Abstract

We study a security threat to batch reinforcement learning and control where the attacker aims to poison the learned policy. The victim is a reinforcement learner / controller which first estimates the dynamics and the rewards from a batch data set, and then solves for the optimal policy with respect to the estimates. The attacker can modify the data set slightly before learning happens, and wants to force the learner into learning a target policy chosen by the attacker. We present a unified framework for solving batch policy poisoning attacks, and instantiate the attack on two standard victims: tabular certainty equivalence learner in reinforcement learning and linear quadratic regulator in control. We show that both instantiation result in a convex optimization problem on which global optimality is guaranteed, and provide analysis on attack feasibility and attack cost. Experiments show the effectiveness of policy poisoning attacks.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyma2019policy

Related papers

• Policy Teaching In Reinforcement Learning Via Environment Poisoning Attacks (2020)0.00
• Online Poisoning Attack Against Reinforcement Learning Under Black-box Environments (2024)0.00
• Reward Poisoning In Reinforcement Learning: Attacks Against Unknown Learners In Unknown Environments (2021)0.00
• Understanding The Limits Of Poisoning Attacks In Episodic Reinforcement Learning (2022)3.58
• Reward Poisoning Attacks On Offline Multi-agent Reinforcement Learning (2022)0.00
• Universal Black-box Reward Poisoning Attack Against Offline Reinforcement Learning (2024)0.00
• Black-box Targeted Reward Poisoning Attack Against Online Deep Reinforcement Learning (2023)0.00
• Execute Order 66: Targeted Data Poisoning For Reinforcement Learning (2022)0.00