Potential-based online policy iteration algorithms for Markov decision processes.
Haitao FangXi-Ren CaoPublished in: IEEE Trans. Autom. Control. (2004)
Keyphrases
- policy iteration
- markov decision processes
- model free
- optimal policy
- finite state
- sample path
- factored mdps
- least squares
- reinforcement learning
- markov decision process
- average reward
- approximate dynamic programming
- policy evaluation
- transition matrices
- state space
- dynamic programming
- average cost
- finite horizon
- temporal difference
- fixed point
- infinite horizon
- reinforcement learning algorithms
- markov decision problems
- markov games
- decision processes
- function approximation
- convergence rate
- linear programming
- actor critic