The Representational Capacity of Action-Value Networks for Multi-Agent Reinforcement Learning

Conference paper (2019)

Authors

Jacopo Castellini University of Liverpool
F.A. Oliehoek Interactive Intelligence - EEMCS
Rahul Savani University of Liverpool
Shimon Whiteson University of Oxford
Research Group
Interactive Intelligence (EEMCS) (TU Delft)
Copyright: © 2019 Jacopo Castellini, F.A. Oliehoek, Rahul Savani, Shimon Whiteson
Neural networks Decision-making Multi-agent systems One-shot games Actionvalue representation
More Info
expand_more

Published Date

2019

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Faculty

Electrical Engineering, Mathematics and Computer Science

Department

Intelligent Systems

Research Group

Interactive Intelligence

Abstract

Recent years have seen the application of deep reinforcement learning techniques to cooperative multi-agent systems, with great empirical success. In this work, we empirically investigate the representational power of various network architectures on a series of one-shot games. Despite their simplicity, these games capture many of the crucial problems that arise in the multi-agent setting, such as an exponential number of joint actions or the lack of an explicit coordination mechanism. Our results quantify how well various approaches can represent the requisite value functions, and help us identify issues that can impede good performance.

Files

TRC.pdf
(.pdf | 1.21 Mb)