September 23, 2014
Grant Will Fund Studies in the Physics Behind Decision-Making
Nov. 15, 2013
Dr. Alain Bensoussan
How are your decisions influenced by those of other people? And how do your decisions affect others?
An emerging area of research aims to answer questions about individual decision-making within very large populations. Called mean field game theory, the research seeks to explain, model and ultimately make predictions about behavior.
Dr. Alain Bensoussan, an Ashbel Smith Professor in the Naveen Jindal School of Management, has received a $339,570, three-year grant from the National Science Foundation for a research project to study a theory called “Mean Field Games, Mean Field Type Control and Extensions.”
Bensoussan, director of the Jindal School’s International Center for Decision and Risk Analysis, said he is excited about the opportunity for UT Dallas to make a contribution to the new field. Much of the research conducted so far has been in Europe.
Mean field game theory applies a concept from physics to game theory – the study of strategic decision-making that involves analyzing how people make decisions in various situations.
Mean field game theory moves the analysis to a larger platform. The term mean field refers to a physics concept that attempts to describe the effect of an infinite number of particles on the motion of a single particle. Researchers began to apply the concept to social sciences in the last decade to study how an infinite number of factors affect individual decisions.
“Our society, our community is like a lot of particles and the interaction is creating something which has to be understood.”
Part of Bensoussan’s research will focus on applying the theory to decision-making related to uncertainty and risk management.
“If you want to analyze how the large community you are imbedded in will influence your behavior, it becomes extremely complex,” he said.
“You do not decide in a desert. You decide in front of other people who also decide,” Bensoussan said. “Our society, our community is like a lot of particles and the interaction is creating something which has to be understood.”
Researchers have used the theory to model “the wave” at packed sports stadiums. Will you briefly stand, yell and raise your arms with everyone when it’s your section’s turn? Will you remain seated or go to the hot dog stand instead?
Mean field game theory has the potential for many applications, such as predicting traffic outcomes to prevent accidents.
“If you are on a highway, you have busy, heavy traffic and you are not going to drive in the same way. You could go faster. Of course, the other drivers influence you and influence your actions,” Bensoussan said.
The theory also could help understand the impact of new technology. Bensoussan gave the example of cloud computing. He said that a decision to use cloud computing can be influenced by many factors, such as whether others use it to the point that it becomes a business necessity.
The theory also may help explain economic phenomena, Bensoussan said. The long-accepted equilibrium theory holds that the market forces of supply and demand bring an economy to a point of equilibrium. But Bensoussan said that equilibrium theory fails to explain the recent economic crisis.
“We need new theory. If mean field can provide something in that direction, it would be useful. If it can provide a good way to study traffic and then prevent accidents and also provide indications to drivers on what to do to drive safely and efficiently, that would be extremely useful.
“If we can understand how cloud technology is going to evolve, the decision-maker who has to decide whether the company will go with the cloud or not will have an easier way to decide," he said. "We will have more informed-based decisions than purely intuition. That’s the potential benefit.”
But first, the theory needs more development, Bensoussan said.
His project, “Mean Field Games, Mean Field Type Control and Extensions,” is recapped in a mathematics brief from international academic publisher Springer. The book, Mean Field Games and Mean Field Type Control Theory, by Bensoussan and co-authors Jens Frehse of the University of Bonn and Philip Yam of the Chinese University of Hong Kong, has just been released.