Revisiting Multi-Objective MDPs with Relaxed Lexicographic Preferences.
Luis Enrique PinedaKyle Hollins WrayShlomo ZilbersteinPublished in: AAAI Fall Symposia (2015)
Keyphrases
- multi objective
- multiple objectives
- markov decision processes
- preference models
- multiple criteria
- evolutionary algorithm
- optimization algorithm
- multi objective optimization
- reinforcement learning
- particle swarm optimization
- objective function
- conflicting objectives
- multi attribute
- state space
- planning under uncertainty
- nsga ii
- user preferences
- optimal policy
- finite horizon
- probabilistic planning
- multi objective optimization problems
- optimal solution
- genetic algorithm
- optimum design
- factored mdps
- multicriteria optimization
- multi objective evolutionary
- bi objective
- pareto optimal
- differential evolution
- decision making
- average cost
- soft constraints
- multi criteria
- combinatorial optimization
- dynamic programming
- learning algorithm