RL STaR Platform: Reinforcement Learning for Simulation based Training of Robots.
Tamir BlumGabin PailletMickael LaineKazuya YoshidaPublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- function approximation
- real robot
- model free
- state space
- supervised learning
- robotic control
- reinforcement learning algorithms
- optimal policy
- multi agent
- cooperative
- markov decision processes
- temporal difference
- transfer learning
- mobile robot
- learning algorithm
- optimal control
- rl algorithms
- robot control
- autonomous learning
- state and action spaces
- action space
- continuous state
- training set
- learning classifier systems
- control problems
- multi robot
- direct policy search
- dynamic programming
- machine learning
- function approximators
- partially observable domains
- actor critic
- solve complex tasks
- learning agent
- policy iteration
- learning capabilities
- human robot interaction
- autonomous robots
- robotic systems
- learning process