Timezone: »

 
Poster
Agent Modelling under Partial Observability for Deep Reinforcement Learning
Georgios Papoudakis · Filippos Christianos · Stefano Albrecht

Thu Dec 09 12:30 AM -- 02:00 AM (PST) @
in Poster Session 5 »

Modelling the behaviours of other agents is essential for understanding how agents interact and making effective decisions. Existing methods for agent modelling commonly assume knowledge of the local observations and chosen actions of the modelled agents during execution. To eliminate this assumption, we extract representations from the local information of the controlled agent using encoder-decoder architectures. Using the observations and actions of the modelled agents during training, our models learn to extract representations about the modelled agents conditioned only on the local observations of the controlled agent. The representations are used to augment the controlled agent's decision policy which is trained via deep reinforcement learning; thus, during execution, the policy does not require access to other agents' information. We provide a comprehensive evaluation and ablations studies in cooperative, competitive and mixed multi-agent environments, showing that our method achieves significantly higher returns than baseline methods which do not use the learned representations.

Keywords: Reinforcement Learning and Planning

Author Information

Georgios Papoudakis (University of Edinburgh)
Filippos Christianos (University of Edinburgh)
Stefano Albrecht (University of Edinburgh)

More from the Same Authors