On Markov Decision Processes with Pseudo-Boolean Reward Functions.
Michael N. KatehakisPublished in: ISAIM (2018)
Keyphrases
- reward function
- markov decision processes
- optimal policy
- integer linear programming
- reinforcement learning algorithms
- state space
- reinforcement learning
- finite state
- partially observable
- dynamic programming
- policy iteration
- planning under uncertainty
- markov decision process
- factored mdps
- average reward
- multiple agents
- infinite horizon
- action space
- combinatorial problems
- linear programming
- decision makers
- least squares