Acceleration Operators in the Value Iteration Algorithms for Markov Decision Processes.
Oleksandr ShlakhterChi-Guhn LeeDmitry KhmelevNasser JaberPublished in: Oper. Res. (2010)
Keyphrases
- markov decision processes
- policy iteration
- factored mdps
- state space
- optimal policy
- reachability analysis
- finite state
- stochastic shortest path
- dynamic programming
- average reward
- decision theoretic planning
- reinforcement learning
- partially observable markov decision processes
- reinforcement learning algorithms
- infinite horizon
- model based reinforcement learning
- planning under uncertainty
- policy evaluation
- learning algorithm
- transition matrices
- computational complexity
- least squares
- continuous state spaces
- finite horizon
- long run
- average cost
- risk sensitive
- optimal control
- sufficient conditions
- markov decision process
- partially observable