Regularized Policy Iteration with Nonparametric Function Spaces.
Amir-massoud FarahmandMohammad GhavamzadehCsaba SzepesváriShie MannorPublished in: J. Mach. Learn. Res. (2016)
Keyphrases
- policy iteration
- bellman residual
- least squares
- sample path
- markov decision processes
- linear approximation
- fixed point
- reinforcement learning
- model free
- optimal policy
- infinite horizon
- asymptotic analysis
- markov decision process
- dynamic programming
- control system
- optimal control
- finite state
- markov chain
- bayesian networks
- temporal difference
- linear program
- state space
- linear programming
- sufficient conditions