Softmax exploration strategies for multiobjective reinforcement learning.
Peter VamplewRichard DazeleyCameron FoalePublished in: Neurocomputing (2017)
Keyphrases
- multi objective
- reinforcement learning
- exploration strategy
- action selection
- temporal difference learning
- evolutionary algorithm
- optimization algorithm
- multiobjective optimization
- multi objective optimization
- active exploration
- particle swarm optimization
- search strategies
- reinforcement learning algorithms
- unknown environments
- genetic algorithm
- machine learning
- function approximation
- artificial bee colony
- multiple objectives
- autonomous learning
- multiobjective genetic algorithm
- model free
- pareto optimal
- nsga ii
- markov decision processes
- conflicting objectives
- state space
- trade off
- multiobjective evolutionary algorithm
- exploration exploitation
- multiobjective evolutionary algorithms
- objective function
- genetic programming
- learning agents
- simulated annealing
- learning process
- multi agent