Value function interference and greedy action selection in value-based multi-objective reinforcement learning.
Peter VamplewCameron FoaleRichard DazeleyPublished in: CoRR (2024)
Keyphrases
- action selection
- reinforcement learning
- multi objective
- temporal difference
- basal ganglia
- robot soccer
- optimization algorithm
- function approximators
- evolutionary algorithm
- decision making
- multi objective optimization
- human robot
- continuous state and action spaces
- multiple objectives
- genetic algorithm
- multi objective optimization problems
- learning algorithm
- state space
- greedy algorithm
- optimal control
- search algorithm
- control policy
- real world
- dynamic programming
- reinforcement learning algorithms
- monte carlo
- particle swarm optimization
- search space
- total reward
- neural network
- action selection mechanism