Malicious Attacks against Deep Reinforcement Learning Interpretations.
Mengdi HuaiJianhui SunRenqin CaiLiuyi YaoAidong ZhangPublished in: KDD (2020)
Keyphrases
- malicious attacks
- reinforcement learning
- function approximation
- model free
- markov decision processes
- reinforcement learning algorithms
- robotic control
- multi agent
- state space
- machine learning
- continuous state
- control problems
- direct policy search
- reinforcement learning methods
- function approximators
- temporal difference
- optimal policy
- semi supervised
- learning process
- learning algorithm
- action selection
- reward function
- active learning
- robot control
- artificial intelligence
- genetic algorithm
- real time
- database