Multi-agent Learning Dynamics: A Survey

H. Jaap van den Herik, Daniel Hennes, Michael Kaisers, Karl Tuyls, Katja Verbeeck

Abstract

In this paper we compare state-of-the-art multi-agent reinforcement learning algorithms in a wide variety of games. We consider two types of algorithms: value iteration and policy iteration. Four characteristics are studied: initial conditions, parameter settings, convergence speed, and local versus global convergence. Global convergence is still difficult to achieve in practice, despite existing theoretical guarantees. Multiple visualizations are included to provide a comprehensive insight into the learning dynamics.
BibTeX
@inproceedings{herik2007multi,
  author = {H. Jaap van den Herik and Daniel Hennes and Michael Kaisers and Karl Tuyls and Katja Verbeeck},
  title = {{Multi-agent Learning Dynamics: A Survey}},
  booktitle = {Cia},
  year = {2007},
  doi = {10.1007/978-3-540-75119-9\_4},
}

Loading related content...

Figure