MORAL: Aligning AI with Human Norms through Multi-Objective Reinforced Active Learning.
Markus PeschlArkady ZgonnikovFrans A. OliehoekLuciano Cavalcante SiebertPublished in: CoRR (2022)
Keyphrases
- multi objective
- active learning
- multi objective optimization
- evolutionary algorithm
- learning strategies
- artificial intelligence
- optimization algorithm
- semi supervised
- genetic algorithm
- artificially intelligent
- expert systems
- ai systems
- conflicting objectives
- human intelligence
- multiobjective optimization
- random sampling
- human behavior
- knowledge based systems
- image registration
- machine learning
- particle swarm optimization
- active learning strategies
- computational intelligence
- case based reasoning
- multiple objectives
- intelligent systems
- multi objective optimization problems
- multi agent systems
- objective function
- machine intelligence
- multi objective evolutionary
- optimum design
- human level
- trade off
- knowledge representation
- pareto optimal
- cost sensitive
- optimization problems
- data sets