Gradient-augmented Supervised Learning of Optimal Feedback Laws Using State-dependent Riccati Equations.
Giacomo AlbiSara BicegoDante KalisePublished in: CoRR (2021)
Keyphrases
- state dependent
- supervised learning
- asymptotically optimal
- optimal policy
- steady state
- queueing networks
- arrival rate
- reinforcement learning
- differential equations
- markov chain
- training data
- product form
- stationary distribution
- service rates
- queue length
- customer demand
- dynamic programming
- inventory level
- optimal production
- optimal control