Podracer architectures for scalable Reinforcement Learning.
Matteo HesselManuel KroissAidan ClarkIurii KemaevJohn QuanThomas KeckFabio ViolaHado van HasseltPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- function approximation
- model free
- optimal policy
- highly scalable
- learning process
- supervised learning
- markov decision processes
- direct policy search
- multi agent reinforcement learning
- parallel architectures
- real robot
- learning capabilities
- reinforcement learning algorithms
- temporal difference
- state space
- dynamic programming
- multi agent
- artificial intelligence
- transfer learning
- mobile robot
- control system
- control problems
- image sequences
- function approximators
- temporal difference learning
- information retrieval
- policy search
- data sets