Parallelizing Parallel Rollout Algorithm for Solving Markov Decision Processes.
Seon Wook KimHyeong Soo ChangPublished in: WOMPAT (2003)
Keyphrases
- markov decision processes
- dynamic programming
- model based reinforcement learning
- learning algorithm
- np hard
- average reward
- reinforcement learning
- stochastic shortest path
- policy iteration
- monte carlo
- computational complexity
- optimality criterion
- objective function
- optimal control
- finite state
- reinforcement learning algorithms
- optimal policy
- probabilistic planning
- transition matrices
- real time dynamic programming