반응형
Main Paper https://ieeexplore.ieee.org/abstract/document/9164559
A survey on multi‐agent deep reinforcement learning: from the perspective of challenges and applications
abstract
세상에는 다양한 multi-agent system이 존재하고, 이러한 system은 task를 수행하기 위해 cooperate or competitive하다. single RL과 비교하여, multi-agent RL은 larger search space, perception of other agents, system robustness를 지닌다. 이 논문은 current multi-agent RL strategy training method와 multi-agent RL의 latest progress를 살펴본다. 마지막으로 multi-agent RL의 application prospects 및 development trends, collabotraion or competition을 소개한다.
Introduction
MAS:
반응형
'논문 리뷰 > MARL algorithm' 카테고리의 다른 글
| [CommNet] Learning Multi-agent Communication with Backpropagation (0) | 2022.07.29 |
|---|---|
| [RIAL & DIAL] Learning to Communicate with Deep Multi-Agent Reinforcement Learning (0) | 2022.07.26 |
| [Distributed W-Learning] Multi-Policy Optimization in Self-Organizing Systems (0) | 2022.07.20 |
| CommNet (learning communication, PG) (0) | 2021.07.12 |
| RIAL & DIAL (learning communication, VB) (0) | 2021.07.07 |
| Multi-agent Intro (0) | 2021.07.06 |
댓글