How to Dynamically Merge Markov Decision Processes.
Satinder P. SinghDavid CohnPublished in: NIPS (1997)
Keyphrases
- markov decision processes
- state space
- policy iteration
- finite state
- optimal policy
- reinforcement learning
- reachability analysis
- dynamic programming
- factored mdps
- transition matrices
- decision processes
- reinforcement learning algorithms
- average cost
- average reward
- finite horizon
- markov decision process
- model based reinforcement learning
- risk sensitive
- state and action spaces
- reward function
- planning under uncertainty
- decision theoretic planning
- data mining
- linear programming
- state abstraction
- infinite horizon
- partially observable
- search algorithm