Generalizing from a few environments in safety-critical reinforcement learning.
Zachary KentonAngelos FilosOwain EvansYarin GalPublished in: CoRR (2019)
Keyphrases
- safety critical
- reinforcement learning
- fault tolerant
- formal methods
- safety analysis
- nuclear power plant
- embedded systems
- agent architecture
- adaptive systems
- regulatory requirements
- state space
- learning algorithm
- formal specification
- support systems
- metamodel
- case based reasoning
- information systems
- machine learning