Web SAE
SAE News SAE News SAE Papoulias receives award-wi...

SAE

Send | Print | RSS SAE '
Papoulias receives award-winning scientist
Athens15.07.2009

President of the Republic Karolos Papoulias on Tuesday received 28 – year – old computer algorithm researcher Dr. Constantine Daskalakis, a Greek academic currently working for MIT, who was recently awarded for a doctoral dissertation that essentially questions whether Nash equilibria are an accurate description of interactions between individuals in all cases.

In his paper, Daskalakis provided new insights into some of the dominant economic theories that have been based on the concepts developed by brilliant but unstable mathematician John Nash in the mid-20th century, as well as other complex systems.

“You represent another Greece and this Greece is very strong,” Papoulias said, congratulating Dr. Daskalakis and his teacher at the University of California, Berkeley, Prof. Christos Papadimitriou that accompanied him.

As Papadimitriou stressed during the meeting, “many theories based on equilibria in economics must now be changed in order for there to be realistic solutions. Today, in the age of the Internet, we must think and introduce new calculations,” he said.

Daskalakis won the 2008 Doctoral Dissertation Award from the Association for Computing Machinery (ACM) for his work on computing Nash equilibria.

In this, he examines whether rational, self-interested individuals can arrive, through their interactions, at a state where no single one of them would be better off switching strategies unless others did so as well. Such a state is called a Nash equilibrium, in honor of John Nash, who defined it, and is traditionally used in Game Theory as a rigorous way of predicting the behavior of people in conflict situations.

Daskalakis showed that in complex systems the Nash equilibrium is computationally unachievable in some cases. This result suggests that the Nash equilibrium may not be an accurate prediction of behavior in all situations and emphasizes the need for new, computationally meaningful methods for modeling strategic behavior in complex systems such as those encountered in financial markets, online systems, and social networks.

Source: ANA-MPA

 
Share |