Potential-based multiobjective reinforcement learning approaches to low-impact agents for AI safety.
Peter VamplewCameron FoaleRichard DazeleyAdam BignoldPublished in: Eng. Appl. Artif. Intell. (2021)
Keyphrases
- multi objective
- reinforcement learning
- multi agent
- multi agent systems
- artificial intelligence
- machine learning
- learning agents
- cooperative
- multiple agents
- autonomous agents
- multiagent systems
- expert systems
- multiagent learning
- intelligent behavior
- trade off
- multi agent reinforcement learning
- multiobjective evolutionary algorithms
- multiobjective genetic algorithm
- multiobjective optimization
- mobile agents
- multiple objectives
- multi objective optimization
- software agents
- resource allocation
- optimization algorithm
- dynamic environments
- case based reasoning
- decision making
- genetic algorithm