Learning Abstractions via Reinforcement Learning

Typ
Examensarbete för masterexamen
Program
Publicerad
2022
Författare
JERGÉUS, ERIK
KARLSSON OINONEN, LEO
Modellbyggare
Tidskriftstitel
ISSN
Volymtitel
Utgivare
Sammanfattning
In this paper we take the first steps in studying a new approach to synthesis of efficient communication schemes in multi-agent systems, trained via reinforcement learning. We combine symbolic methods with machine learning, in what is referred to as a neuro-symbolic system. The agents are not restricted to only use initial primitives: reinforcement learning is interleaved with steps to extend the current language with novel higher-level concepts, allowing generalisation and more informative communication via shorter messages. We demonstrate that this approach allow agents to converge more quickly on a small collaborative construction task.
Beskrivning
Ämne/nyckelord
RL, MARL, multi-agent, DreamCoder, neuro-symbolic, abstraction, communication, AI
Citation
Arkitekt (konstruktör)
Geografisk plats
Byggnad (typ)
Byggår
Modelltyp
Skala
Teknik / material