Your Room is not Private: Gradient Inversion Attack for Deep Q-Learning.
Miao LiWenhao DingDing ZhaoPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- state space
- cooperative
- function approximation
- learning algorithm
- multi agent
- edge detection
- optimal policy
- privacy preserving
- image reconstruction
- stochastic approximation
- private data
- deep learning
- action selection
- countermeasures
- multi agent reinforcement learning
- learning rate
- image gradient
- private information
- potential field
- dynamic programming
- reinforcement learning algorithms
- neural network
- gradient method
- temporal difference learning
- gradient direction
- abstract argumentation
- malicious users