Integral Policy Iteration for Zero-Sum Games with Completely Unknown Nonlinear Dynamics.
Hongliang LiDerong LiuDing WangPublished in: ICONIP (1) (2013)
Keyphrases
- nonlinear dynamics
- policy iteration
- markov decision processes
- fixed point
- model free
- optimal policy
- least squares
- sample path
- reinforcement learning
- neural network
- dynamical systems
- markov decision process
- temporal difference
- average reward
- optimal control
- finite state
- convergence rate
- policy evaluation
- optimal strategy
- linear programming
- infinite horizon
- state space
- function approximation
- dynamic programming
- imperfect information
- markov decision problems
- asymptotic analysis
- optimal solution
- markov chain