GPU Parallelization of Policy Iteration RRT#.
R. Connor LawsonLinda WillsPanagiotis TsiotrasPublished in: IROS (2020)
Keyphrases
- policy iteration
- parallel processing
- markov decision processes
- model free
- fixed point
- optimal policy
- reinforcement learning
- least squares
- sample path
- markov decision process
- finite state
- real time
- policy evaluation
- linear programming
- state space
- infinite horizon
- temporal difference
- optimal control
- parallel implementation
- graphics processing units
- average reward
- markov decision problems
- dynamic programming
- shared memory
- function approximation
- pairwise
- convergence rate
- multistage