LASER: Learning a Latent Action Space for Efficient Reinforcement Learning.
Arthur AllshireRoberto Martín-MartínCharles LinShawn ManuelSilvio SavareseAnimesh GargPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- learning algorithm
- action space
- reinforcement learning methods
- state space
- action selection
- prior knowledge
- state action
- control problems
- reinforcement learning algorithms
- continuous state
- function approximators
- function approximation
- supervised learning
- multi agent systems
- markov decision process
- machine learning
- real valued
- rl algorithms
- continuous state spaces