Defending Against Label-Only Attacks via Meta-Reinforcement Learning.
Dayong YeTianqing ZhuKun GaoWanlei ZhouPublished in: IEEE Trans. Inf. Forensics Secur. (2024)
Keyphrases
- reinforcement learning
- ddos attacks
- countermeasures
- function approximation
- multi label
- image labeling
- denial of service attacks
- markov decision processes
- machine learning
- dos attacks
- meta level
- class labels
- robotic control
- learning algorithm
- learning process
- multi agent
- state space
- dynamic programming
- neural network
- computer security
- model free
- traffic analysis
- reinforcement learning algorithms
- temporal difference
- malicious attacks
- watermarking algorithm
- multi agent reinforcement learning
- supervised learning
- reinforcement learning methods
- chosen plaintext
- temporal difference learning
- function approximators
- action space
- learning classifier systems
- information security
- optimal policy
- signal processing
- cooperative