Deep Conservative Policy Iteration.
Nino VieillardOlivier PietquinMatthieu GeistPublished in: CoRR (2019)
Keyphrases
- policy iteration
- markov decision processes
- optimal policy
- least squares
- model free
- reinforcement learning
- fixed point
- sample path
- infinite horizon
- policy evaluation
- finite state
- average reward
- temporal difference
- optimal control
- markov decision problems
- markov decision process
- linear programming
- state space
- dynamic programming
- discounted reward
- convergence rate
- image sequences
- reward function
- markov chain
- objective function