A Terminal Condition with A Local Stable Manifold for Trajectory Collections of An Infinite Horizon Optimal Control.
Takafumi OkiShigeo WadaPublished in: ECC (2022)
Keyphrases
- optimal control
- infinite horizon
- dynamic programming
- finite horizon
- single item
- production planning
- stochastic demand
- sufficient conditions
- control strategy
- reinforcement learning
- partially observable
- holding cost
- periodic review
- average cost
- optimal control problems
- fixed cost
- bayesian networks
- lost sales
- infinite dimensional
- machine learning
- markov decision problems
- policy iteration
- actor critic
- markov decision process