Accelerating Deep Continuous Reinforcement Learning through Task Simplification.
Matthias KerzelHadi Beik MohammadiMohammad-Ali ZamaniStefan WermterPublished in: IJCNN (2018)
Keyphrases
- reinforcement learning
- action space
- continuous state spaces
- function approximation
- model free
- state space
- continuous domains
- information retrieval
- learning capabilities
- preprocessing
- policy search
- learning algorithm
- continuous state
- temporal difference learning
- robot control
- machine learning
- robotic control
- reinforcement learning algorithms
- temporal difference
- learning problems
- markov decision processes
- transfer learning
- optimal policy
- supervised learning
- dynamic programming
- learning process
- multi agent
- information systems