Integral Reinforcement Learning for Linear Continuous-Time Zero-Sum Games With Completely Unknown Dynamics.
Hongliang LiDerong LiuDing WangPublished in: IEEE Trans Autom. Sci. Eng. (2014)
Keyphrases
- reinforcement learning
- dynamical systems
- optimal control
- state space
- function approximation
- markov chain
- dynamic programming
- vector autoregressive
- optimal strategy
- dynamic model
- temporal difference
- linear systems
- reinforcement learning algorithms
- stochastic processes
- function approximators
- temporal difference learning
- closed form
- learning algorithm
- linear model
- markov decision processes
- model free
- collective behavior
- probability distribution