An Abstraction-based Method to Check Multi-Agent Deep Reinforcement-Learning Behaviors.
Pierre El MqirmiFrancesco BelardinelliBorja G. LeónPublished in: AAMAS (2021)
Keyphrases
- reinforcement learning
- multi agent
- high accuracy
- experimental evaluation
- computational complexity
- significant improvement
- high precision
- detection method
- clustering method
- computationally efficient
- unsupervised learning
- prior knowledge
- pairwise
- preprocessing
- evolutionary algorithm
- probabilistic model
- semi supervised
- multi agent systems
- edge detection
- feature set
- particle filter
- cooperative
- multiscale
- high level
- machine learning
- neural network