Model-Free Policy Gradients for Multi-Agent Shape Formation

Elizabeth Boroson1, Fei Sha1, Nora Ayanian1

  • 1University of Southern California

Details

Sessions

10:00 - 10:30 | Mon 25 Sep | Ballroom Foyer | MoAmPo

Monday Posters AM

Full Text

Abstract

We develop a model-free technique based on gradient ascent for multi-agent systems to learn distributed policies for tight coordination, and we demonstrate its performance on the problem of shape formation.

Additional Information

No information added

Video

No videos found