Variance-Reduced Conservative Policy Iteration.
Naman AgarwalBrian BullinsKaran SinghPublished in: ALT (2023)
Keyphrases
- policy iteration
- markov decision processes
- model free
- reinforcement learning
- fixed point
- sample path
- temporal difference
- least squares
- optimal policy
- finite state
- linear programming
- state space
- optimal control
- infinite horizon
- markov decision process
- policy evaluation
- multistage
- convergence rate
- dynamic programming
- average reward
- markov decision problems
- markov chain
- cost function
- bayesian networks
- graphical models