Empirical evaluation methods for multiobjective reinforcement learning algorithms.
Peter VamplewRichard DazeleyAdam BerryRustam IssabekovEvan DekkerPublished in: Mach. Learn. (2011)
Keyphrases
- evaluation methods
- reinforcement learning algorithms
- multi objective
- reinforcement learning
- evaluation method
- state space
- markov decision processes
- evolutionary algorithm
- reinforcement learning problems
- model free
- temporal difference
- evaluation criteria
- function approximation
- optimization algorithm
- reinforcement learning methods
- learning algorithm
- objective function
- particle swarm optimization
- genetic algorithm
- evaluation metrics
- machine learning
- evaluation measures
- dynamic environments
- reward function
- pareto optimal
- precision and recall
- user centred evaluation
- state variables
- active learning
- search algorithm