On Finding Compromise Solutions in Multiobjective Markov Decision Processes.
Patrice PernyPaul WengPublished in: ECAI (2010)
Keyphrases
- partially observable
- markov decision processes
- multi objective
- state space
- multiple objectives
- finite state
- reinforcement learning
- multiobjective optimization
- optimal policy
- policy iteration
- infinite horizon
- pareto optimal
- reachability analysis
- dynamic programming
- finite horizon
- evolutionary algorithm
- model based reinforcement learning
- decision theoretic planning
- planning under uncertainty
- decision processes
- multi objective optimization
- reinforcement learning algorithms
- transition matrices
- optimal solution
- markov decision process
- average reward
- risk sensitive
- average cost
- fixed point
- machine learning
- policy evaluation
- state and action spaces
- action sets
- search algorithm