Synchronizing Objectives for Markov Decision Processes
Laurent DoyenThierry MassartMahsa ShirmohammadiPublished in: iWIGP (2011)
Keyphrases
- markov decision processes
- optimal policy
- finite state
- dynamic programming
- transition matrices
- state space
- reinforcement learning
- factored mdps
- decision theoretic planning
- average cost
- multiple objectives
- risk sensitive
- planning under uncertainty
- average reward
- reachability analysis
- reinforcement learning algorithms
- model based reinforcement learning
- policy iteration
- finite horizon
- decision processes
- semi markov decision processes
- state and action spaces
- action sets
- machine learning
- action space
- partially observable
- infinite horizon
- state abstraction
- partially observable markov decision processes
- optimal control