Algorithms for Multi-agent Reinforcement Learning in Complex Environments

复杂环境下多智能体强化学习算法

• 批准号: 2427742
• 金额: --
• 依托单位: University of Edinburgh
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2020
• 资助国家: 英国
• 起止时间: 2020 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2427742
• 关键词: Algorithms; Multi; agent; Reinforcement; Learning

This project will investigate the combination of causal inference and deep reinforcement learning to design an algorithm with the ability to learn causal relationships through exploration in a multi-agent system. The goal is for an agent to apply the learned causal relationships to decision-making and transfer causal knowledge to new environments.

该项目将研究因果推理和深度强化学习的结合，以设计一种能够通过多智能体系统中的探索来学习因果关系的算法。我们的目标是一个代理应用学到的因果关系的决策和因果知识转移到新的环境。

项目成果

Temporal Disentanglement of Representations for Improved Generalisation in Reinforcement Learning

• DOI: 10.48550/arxiv.2207.05480
• 发表时间: 2022-07
• 期刊: ArXiv
• 作者: Mhairi Dunion;Trevor A. McInroe;K. Luck;Josiah P. Hanna;Stefano V. Albrecht