Evaluation of strategies for fitness-driven random walkers on a graph
From Santa Fe Institute Events Wiki
Evolutionary dynamics of fitness-driven walkers on a graph
Roberta Sinatra*, Erika Fille Legara, and Chaitanya Gokhale
In this project, we introduce a model that investigates how strategies of interacting individuals in a group evolve as they are allowed to interact with other agents and migrate to other neighboring groups through a biased random walk. Here, we construct varying complex network models to create different grouping topologies. Each node in a network represents a group of players in a well mixed state. Agent dynamics is governed by game theoretic rules where agents in a group/class interact with each other playing either one of two strategies, A or B.
Population of players using strategies A and B are then evaluated resulting in an average payoff of the whole population in node i. Each agent then migrates to one of its neighboring node j through a biased random walk by evaluating and contrasting the different fitness values of its group's neighbors.
This model is reminiscent of social systems and ecological systems where agents migrate to neighboring environments depending on the fitness or prestige of the said neighbors. In social systems, for example, this may describe how and why individuals would move from one neighboring city to another. This may also tell of why employees of a certain company would want to move to other companies.
(Details to be uploaded soon.)