Global Convergence of Policy Gradient for Sequential Zero-Sum Linear Quadratic Dynamic Games.
Jingjing BuLillian J. RatliffMehran MesbahiPublished in: CoRR (2019)
Keyphrases
- global convergence
- policy gradient
- game theory
- linear quadratic
- optimization methods
- global optimum
- stochastic games
- nash equilibria
- optimal control
- convergence speed
- convergence rate
- incomplete information
- dynamic environments
- neural network
- closed loop
- optimization method
- differential evolution
- reinforcement learning algorithms
- model selection
- average reward
- optimization problems
- reinforcement learning
- machine learning