Inductive Policy Selection for First-Order MDPs.
Sung Wook YoonAlan FernRobert GivanPublished in: UAI (2002)
Keyphrases
- optimal policy
- markov decision processes
- decision diagrams
- markov decision process
- policy iteration
- state space
- finite horizon
- policy search
- markov decision problems
- average reward
- reinforcement learning
- state and action spaces
- average cost
- infinite horizon
- finite state
- first order logic
- knowledge representation
- reinforcement learning problems
- reward function
- inductive inference
- decision processes
- dynamic programming
- inductive learning
- planning under uncertainty
- higher order
- factored mdps