Safe Reinforcement Learning in Black-Box Environments via Adaptive Shielding.
Daniel BethellSimos GerasimouRadu CalinescuCalum ImriePublished in: CoRR (2024)
Keyphrases
- black box
- reinforcement learning
- black boxes
- white box
- rule extraction
- hybrid systems
- test cases
- integration testing
- markov decision processes
- reinforcement learning algorithms
- adaptive control
- adaptive learning
- white box testing
- neural network
- function approximation
- optimal policy
- machine learning
- model free
- state transition
- learning capabilities
- intelligent systems
- source code
- hidden markov models