Overlapping Neighborhoods, Local Interaction Model
Open Access
Author:
Levi, Max Richard
Area of Honors:
Mathematics
Degree:
Bachelor of Science
Document Type:
Thesis
Thesis Supervisors:
Dr. Christopher Byrne, Thesis Supervisor Dr. Diane Marie Henderson, Thesis Honors Advisor Dr. Christopher Byrne, Thesis Supervisor
Abstract:
The research develops an overlapping neighborhoods local interaction model in game theory. The agents of the game interact by playing variants of the prisoner’s dilemma. The interaction model represents a social network--Facebook or a suburban neighborhood, for example. Agents of the game are characterized by a simple adaptive learning algorithm, in which strategies are periodically switched to mimic the most successful players.
Past research on this topic began with examining 1-dimensional and 2-dimensional social network models. This research develops past investigations by making a more intuitive model of the social network, with hopes of investigating the impact of the network on behavior, given the assumptions about interaction and learning rules. Starting with a geodesic dome as a simple 2-dimensional network, different aspects of the social interaction are parameterized.
This research provides the framework for the future development of computer simulations that will be used to explore long run model behavior.