Opinion Dynamics for Decision Making and Learning in Multi-agent Interactions



I will discuss a recently proposed model of opinion dynamics and its use in decision making and learning for multi-agent cooperative and strategic interactions. The model describes continuous-time opinion dynamics for an arbitrary number of agents that communicate over a network and form real-valued opinions on an arbitrary number of options about which they may receive payoffs or input from the environment. Many models in the literature update agent opinions using a weighted average of their neighbors’ opinions. This model instead applies a sigmoidal saturation function to opinion exchanges. The saturation of distributed information provides a naturally smooth limit on the influence of neighbors. It also makes the update fundamentally nonlinear: opinions form through a bifurcation yielding multi-stability of network opinion configurations. Leveraging idealized symmetry in the system, qualitative behavioral regimes can be distinguished explicitly in terms of just a few parameters. I will show how to apply the opinion dynamics to design of efficient dynamic multi-agent task allocation and as a model of reinforcement learning in multi-agent finite games.

This is joint work with Anastasia Bizyaeva, Alessio Franci, and Shinkyu Park and based on papers https://arxiv.org/abs/2009.04332v2, https://arxiv.org/abs/2103.14764, and https://arxiv.org/abs/2103.12223.


Naomi Ehrich Leonard is Edwin S. Wilsey Professor of Mechanical and Aerospace Engineering and associated faculty in Applied and Computational Mathematics at Princeton University.  She is also Director of Princeton’s Council on Science and Technology and Editor of the Annual Review of Control, Robotics, and Autonomous Systems. She received her BSE in Mechanical Engineering from Princeton University and her PhD in Electrical Engineering from the University of Maryland.  She is a MacArthur Fellow, elected member of the American Academy of Arts and Sciences, and Fellow of SIAM, IEEE, IFAC, and ASME.  Her current research focuses on dynamics and control of multi-agent systems on networks with application to distributed decision making, spreading processes, multi-robot teams, and collective behavior.

  • CPS PI Meeting 2021
  • Keynote - Naomi Leonard
  • Presentation
Submitted by Naomi Leonard on Wed, 06/09/2021 - 12:53