Reinforcement learning for non-prehensile manipulation: Transfer from simulation to physical system.
Kendall LowreySvetoslav KolevJeremy DaoAravind RajeswaranEmanuel TodorovPublished in: SIMPAR (2018)
Keyphrases
- reinforcement learning
- transfer learning
- multi agent
- reinforcement learning algorithms
- function approximation
- real time
- learning process
- simulation environment
- temporal difference
- mathematical model
- simulation model
- robocup soccer
- simulation models
- knowledge transfer
- simulation study
- markov decision processes
- state space
- evolutionary algorithm
- multi agent systems
- real world
- data sets