Information Relaxation Bounds for Infinite Horizon Markov Decision Processes.
David B. BrownMartin B. HaughPublished in: Oper. Res. (2017)
Keyphrases
- markov decision processes
- infinite horizon
- optimal policy
- finite horizon
- state space
- dynamic programming
- markov decision process
- average cost
- policy iteration
- upper bound
- partially observable
- reinforcement learning
- optimal control
- decision problems
- transition matrices
- lower bound
- multi agent
- decision making
- planning under uncertainty
- probabilistic planning
- decision theoretic planning