Rethinking Adversarial Policies: A Generalized Attack Formulation and Provable Defense in RL.
Xiangyu LiuSouradip ChakrabortyYanchao SunFurong HuangPublished in: ICLR (2024)
Keyphrases
- optimal policy
- reinforcement learning
- multi agent
- defense mechanisms
- ddos attacks
- control policies
- markov decision process
- state space
- multiagent reinforcement learning
- markov decision processes
- intrusion detection
- network security
- countermeasures
- policy search
- control policy
- hierarchical reinforcement learning
- function approximation
- infinite horizon
- decision problems
- multistage
- robust optimization
- reward function
- data mining
- wireless sensor networks
- learning algorithm