Metaheuristics-based Exploration Strategies for Multi-Objective Reinforcement Learning.
Florian FeltenGrégoire DanoyEl-Ghazali TalbiPascal BouvryPublished in: ICAART (2) (2022)
Keyphrases
- multi objective
- reinforcement learning
- exploration strategy
- particle swarm optimization
- genetic algorithm
- optimization algorithm
- multi objective optimization
- evolutionary algorithm
- action selection
- active exploration
- conflicting objectives
- optimization problems
- state space
- multiple objectives
- metaheuristic
- optimal control
- unknown environments
- function approximation
- multiobjective optimization
- search strategies
- learning agents
- vehicle routing problem
- objective function
- bi objective
- tabu search
- model free
- multi agent
- learning algorithm
- simulated annealing
- dynamic programming
- differential evolution
- markov decision processes
- optimal policy
- particle swarm optimization pso
- particle swarm optimisation
- stochastic approximation
- autonomous learning
- machine learning