Do Androids Dream of Electric Fences? Safety-Aware Reinforcement Learning with Latent Shielding.
Peter HeBorja G. LeónFrancesco BelardinelliPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- function approximation
- latent variables
- fuel consumption
- state space
- reinforcement learning algorithms
- temporal difference
- model free
- safety analysis
- genetic algorithm
- learning algorithm
- markov decision processes
- optimal control
- safety critical
- function approximators
- electro mechanical
- robotic control
- neural network
- learning capabilities
- least squares
- hidden markov models
- learning process
- information retrieval
- machine learning