On functional equations for Kth best policies in Markov decision processes.
Hyeong Soo ChangPublished in: Autom. (2013)
Keyphrases
- markov decision processes
- optimal policy
- markov decision process
- reward function
- decision processes
- average cost
- decentralized control
- total reward
- finite state
- state space
- reinforcement learning
- decision problems
- dynamic programming
- policy iteration algorithm
- transition matrices
- discounted reward
- partially observable markov decision processes
- policy iteration
- macro actions
- long run
- infinite horizon
- finite horizon
- factored mdps
- planning under uncertainty
- stationary policies
- expected reward
- decision theoretic planning
- partially observable
- average reward
- reachability analysis
- model based reinforcement learning
- markov decision problems
- state and action spaces
- control policies
- multistage
- heuristic search
- mathematical model
- action space
- sufficient conditions
- least squares