Agent manipulator: Stealthy strategy attacks on deep reinforcement learning.
Jinyin ChenXueke WangYan ZhangHaibin ZhengShanqing YuLiang BaoPublished in: Appl. Intell. (2023)
Keyphrases
- reinforcement learning
- exploration strategy
- multi agent
- action selection
- multi agent systems
- intelligent agents
- learning agent
- multiagent systems
- function approximation
- learning mechanism
- degrees of freedom
- path planning
- computer controlled
- autonomous learning
- state space
- countermeasures
- multiple agents
- reward shaping
- action space
- control policy
- learning capabilities
- reinforcement learning algorithms
- decision making
- agent architecture
- mobile agents
- vision system
- reward function
- model free
- single agent
- real robot
- control method
- autonomous agents
- inverse kinematics
- digital images
- state abstraction
- dynamic programming
- multi agent environments
- machine learning
- robocup soccer