Smooth markets: A basic mechanism for organizing gradient-based learners
Machine Learning
2020-01-22 v2 Artificial Intelligence
Computer Science and Game Theory
Multiagent Systems
Machine Learning
Abstract
With the success of modern machine learning, it is becoming increasingly important to understand and control how learning algorithms interact. Unfortunately, negative results from game theory show there is little hope of understanding or controlling general n-player games. We therefore introduce smooth markets (SM-games), a class of n-player games with pairwise zero sum interactions. SM-games codify a common design pattern in machine learning that includes (some) GANs, adversarial training, and other recent algorithms. We show that SM-games are amenable to analysis and optimization using first-order methods.
Keywords
Cite
@article{arxiv.2001.04678,
title = {Smooth markets: A basic mechanism for organizing gradient-based learners},
author = {David Balduzzi and Wojciech M Czarnecki and Thomas W Anthony and Ian M Gemp and Edward Hughes and Joel Z Leibo and Georgios Piliouras and Thore Graepel},
journal= {arXiv preprint arXiv:2001.04678},
year = {2020}
}
Comments
18 pages, 3 figures