Gradient-augmented Supervised Learning of Optimal Feedback Laws Using State-Dependent Riccati Equations.
Giacomo AlbiSara BicegoDante KalisePublished in: IEEE Control. Syst. Lett. (2022)
Keyphrases
- state dependent
- supervised learning
- steady state
- optimal policy
- asymptotically optimal
- queueing networks
- arrival rate
- service rates
- reinforcement learning
- markov chain
- training data
- queue length
- stationary distribution
- product form
- learning algorithm
- differential equations
- single server
- dynamic programming
- training set
- optimal solution
- bayesian networks
- inventory level
- hamilton jacobi