The Research On Aspiration Driven And Opinion Dynamics Via Spatial Modular Network

In the real world,every individual must make decisions for everything that happens in her whole life,for example,cooperative hunting and migration in the animal world,and communication,interaction,production and life in human society.After selection,the individual will receive the corresponding feedback from the environment she lives in.So everyone will adjust her strategy to fit into the environment.As time goes on,the evolution of behavioral decisions will influence not only the individual but also the whole population.Hence,there are many existing works that focus on behavioral decisions in the field of social science.Based on evolutionary game theory and opinion dynamic model,this work focus on exploring the individual behavior that is impacted by aspiration-driven and opinion-driven on the spatial modular network.The research contents in this work consists of two parts.The first part focus on the evolution of behavior based on aspiration-driven.Based on the full-connected spatial modular networks and the public good game model,this part explores the individual migration behavior and cooperative behavior.The difference between an individual’s gain from gaming and expected gain in the group is used as a driving force for leaving the group.The higher the gain from gaming,the closer the individual’s expected gain,and the individual will not easily leave the current group.Conversely,the individual will join a new group and establish a new network of connected edges.We investigate the evolution of the final population cooperation rate with different expected benefit levels and find that the higher the expected benefit level is,the greater the population cooperation rate is,and the expected benefit level and the gain factor in the public goods game jointly affect the evolution of the population cooperation rate.In second part,the network structure within the community is different from outside the com-munity.We explore the individual migration behavior and cooperative behavior via Q-learning model and opinion threshold model.The results show that individual’s tol-erance threshold influences on the final distribution of population opinion directly.In addition,by comparing different network structures and network sizes,we find that network structure and size have no significant influence on the evolution of population opinion.This work further improves the research on spatial modular networks,especially on dynamic networks,and has a further understanding of individual migration in space.At the same time,we use opinion-driven mechanism to study individual migration be-havior,which provides a new perspective for studying the transfer mechanism and the application of opinion dynamics.In addition,we design an adaptive transfer model based on reinforcement learning,which provides a more objective and effective learn-ing mechanism for simulating individual behavior decisions and helps deepen the inte-gration of the two fields.