How You Act Tells a Lot: Privacy-Leakage Attack on Deep Reinforcement Learning.
Xinlei PanWeiyao WangXiaoshuai ZhangBo LiJinfeng YiDawn SongPublished in: CoRR (2019)
Keyphrases
- reinforcement learning
- information leakage
- personal information
- security risks
- sensitive information
- authentication protocol
- function approximation
- private information
- privacy preserving
- reinforcement learning algorithms
- optimal policy
- third party
- privacy preservation
- optimal control
- differential privacy
- private data
- privacy concerns
- machine learning
- countermeasures
- learning algorithm
- privacy aware
- temporal difference
- state space
- markov decision processes
- multi agent
- privacy issues
- supervised learning
- personal data
- privacy preserving data mining
- model free
- security issues
- data privacy
- statistical databases
- data quality
- learning process
- social networking
- identity management
- privacy guarantees
- confidential information
- information loss