Safe Policy Iteration: A Monotonically Improving Approximate Policy Iteration Approach.
Alberto Maria MetelliMatteo PirottaDaniele CalandrielloMarcello RestelliPublished in: J. Mach. Learn. Res. (2021)
Keyphrases
- policy iteration
- approximate policy iteration
- markov decision processes
- model free
- reinforcement learning
- optimal policy
- fixed point
- least squares
- finite state
- infinite horizon
- temporal difference
- average reward
- markov decision problems
- markov decision process
- convergence rate
- optimal control
- linear programming
- cooperative
- dynamic programming
- reinforcement learning algorithms
- state space
- policy search
- average cost
- dynamical systems
- monte carlo
- neural network