Rode: Learning roles to decompose multi-agent tasks

T Wang, T Gupta, A Mahajan, B Peng… - ar** for monte carlo tree search
A Velasquez, B Bissey, L Barak, A Beckus… - Proceedings of the …, 2021 - ojs.aaai.org
Reinforcement learning and planning have been revolutionized in recent years, due in part
to the mass adoption of deep convolutional neural networks and the resurgence of powerful …

Estimating q (s, s') with deep deterministic dynamics gradients

A Edwards, H Sahni, R Liu, J Hung… - International …, 2020 - proceedings.mlr.press
In this paper, we introduce a novel form of value function, $ Q (s, s') $, that expresses the
utility of transitioning from a state $ s $ to a neighboring state $ s'$ and then acting optimally …

Know your action set: Learning action relations for reinforcement learning

A Jain, N Kosaka, KM Kim, JJ Lim - International Conference on …, 2021 - openreview.net
Intelligent agents can solve tasks in various ways depending on their available set of
actions. However, conventional reinforcement learning (RL) assumes a fixed action set. This …

Learning agent representations for ice hockey

G Liu, O Schulte, P Poupart… - Advances in Neural …, 2020 - proceedings.neurips.cc
Team sports is a new application domain for agent modeling with high real-world impact. A
fundamental challenge for modeling professional players is their large number (over 1K) …

A Transfer Approach Using Graph Neural Networks in Deep Reinforcement Learning

T Yang, H You, J Hao, Y Zheng… - Proceedings of the AAAI …, 2024 - ojs.aaai.org
Transfer learning (TL) has shown great potential to improve Reinforcement Learning (RL)
efficiency by leveraging prior knowledge in new tasks. However, much of the existing TL …

Similarity-based Knowledge Transfer for Cross-Domain Reinforcement Learning

SA Serrano, J Martinez-Carranza, LE Sucar - arxiv preprint arxiv …, 2023 - arxiv.org
Transferring knowledge in cross-domain reinforcement learning is a challenging setting in
which learning is accelerated by reusing knowledge from a task with different observation …