Optimal bounds for numerical approximations of infinite horizon problems based on dynamic programming approach.
Javier de FrutosJulia NovoPublished in: CoRR (2021)
Keyphrases
- infinite horizon
- dynamic programming
- optimal control
- dec pomdps
- finite horizon
- average cost
- optimal policy
- markov decision processes
- state space
- stochastic demand
- production planning
- single item
- markov decision process
- optimal solution
- production capacity
- multistage
- decision problems
- worst case
- fixed cost
- linear programming
- holding cost
- inventory policy
- reinforcement learning
- long run
- upper bound
- lower bound