Action Spaces in Deep Reinforcement Learning to Mimic Human Input Devices.
Marco PleinesFrank ZimmerVincent-Pierre BergesPublished in: CoG (2019)
Keyphrases
- action space
- reinforcement learning
- state space
- continuous state
- markov decision processes
- state and action spaces
- input device
- real valued
- function approximation
- stochastic processes
- continuous state spaces
- action selection
- function approximators
- single agent
- markov decision process
- learning algorithm
- machine learning
- state action
- average reward
- dynamic programming
- multi agent