Tunable and generic problem instance generation for multi-objective reinforcement learning.
Deon GarrettJordi BiegerKristinn R. ThórissonPublished in: ADPRL (2014)
Keyphrases
- multi objective
- reinforcement learning
- evolutionary algorithm
- multi objective optimization
- optimization algorithm
- multi objective optimization problems
- evolutionary optimization
- state space
- genetic algorithm
- multiple objectives
- nsga ii
- multiobjective optimization
- pareto optimal
- model free
- generation process
- conflicting objectives
- temporal difference
- multi objective evolutionary
- particle swarm optimization
- reinforcement learning algorithms
- data sets
- generation method
- high level
- neural network
- multiagent evolutionary algorithm
- robotic control
- optimum design
- markov decision process
- multi criteria
- optimal policy
- domain specific