Myopic Bounds for Optimal Policy of POMDPs: An Extension of Lovejoy's Structural Results.
Vikram KrishnamurthyUdit PareekPublished in: Oper. Res. (2015)
Keyphrases
- optimal policy
- infinite horizon
- partially observable markov decision processes
- markov decision processes
- dynamic programming
- reinforcement learning
- decision problems
- finite horizon
- partially observable
- state space
- finite state
- asymptotically optimal
- long run
- state dependent
- multistage
- markov decision problems
- policy iteration
- lower bound
- markov decision process
- average reward
- bayesian reinforcement learning
- average cost
- upper bound
- initial state
- production planning
- large deviations
- expected reward
- demand distributions
- sufficient conditions
- serial inventory systems
- upper and lower bounds
- optimal control
- inventory control
- belief state
- dynamical systems
- supply chain
- search algorithm