Parallel Q-Learning: Scaling Off-policy Reinforcement Learning under Massively Parallel Simulation.
Zechu LiTao ChenZhang-Wei HongAnurag AjayPulkit AgrawalPublished in: ICML (2023)
Keyphrases
- massively parallel
- reinforcement learning
- parallel computers
- function approximation
- parallel computing
- fine grained
- reinforcement learning algorithms
- state space
- high performance computing
- real robot
- model free
- action selection
- temporal difference learning
- learning algorithm
- parallel machines
- optimal policy
- reinforcement learning methods
- parallel programming
- multi agent
- parallel architectures
- processing elements
- message passing interface
- policy iteration
- dynamic programming
- temporal difference
- multi agent reinforcement learning
- mesh connected
- continuous state and action spaces
- markov decision processes
- parallel execution
- continuous state spaces
- relational reinforcement learning