【作者】 李燕；

【导师】 汪丁丁； 叶航；

【作者基本信息】 浙江大学 ， 西方经济学， 2017， 博士

【摘要】 社会何以可能是社会科学的基本问题,而合作何以可能是行为经济学的基本问题。人与人之间的合作,是人类文明社会的基础,普遍存在于社会生活中。但从进化论的角度看,为什么这种带有利他主义的行为能够保留下来,仍然是一个未解之谜。基于社会网络的个体迁徙在回答这个问题上起到了重要的作用。社会网络上的个体迁徙可以促使合作行为的聚集,它是实现社会网络结构与合作共生演化的重要途径之一。同时,计算经济学的方法为解释复杂系统中合作行为的涌现提供了有利的工具。本文在此背景下利用计算经济学的方法探讨不同迁徙机制对空间囚徒困境博弈中合作解的影响。本文从网络结构和个体迁徙的视角出发,以演化博弈论作为理论工具,以基于Agent的计算经济学作为研究工具,探讨了空间囚徒困境博弈中合作解的演化。具体来说,我们通过基于Netlogo的社会仿真的研究方法,在空间囚徒困境博弈的演化模型中分别探讨了“近君子,远小人”的迁徙策略、基于异质性风险偏好的个体迁徙、考虑社会福利的个体迁徙对合作演化的影响,以此为出发点回答了社会网络上的个体迁徙是如何影响合作行为的演化的。首先,我们研究了“近君子,远小人”的迁徙机制对合作演化的影响。“近君子,远小人”的迁徒机制是指,当个体的邻居全部是合作者时个体不迁徙,这是“近君子”的行为；只要个体的邻居有背叛者时,个体就会以一定的概率进行迁移,这是“远小人”的行为。为了给出“远小人”的概率,我们建立了基于成本与收益的迁徙模型,以此来研究迁徙成本对群体合作水平的作用。仿真结果表明：在较广的参数空间内,“近君子,远小人”的个体迁徙提高了群体的合作水平,为我们理解社会中的合作行为提供了新思路。为了更直观的理解空间中合作团块的涌现,我们展示了随时间演化的空间状态子图。另外,为了探讨此机制在其他网络结构上的有效性,我们研究了小世界网络上的合作演化。结果显示,在小世界网络上,“近君子,远小人”的迁徙模型对合作行为的促进作用仍然有效。接着,我们探究了基于风险偏好的个体迁徙对空间囚徒困境博弈中合作解的影响。以往的研究中要么不考虑风险偏好,要么假定个体在迁徙的过程中是风险中性的,然而,在现实生活中,个体的风险偏好类型是多样的,是异质性的。为此,我们建立了基于风险偏好的个体迁徒模型,通过计算机仿真模拟了行为群体的动态演化过程,探讨了同质性风险偏好和异质性风险偏好对合作行为影响的区别。研究结果表明,当假定群体是同质性风险偏好时,风险厌恶型的个体不支持合作的建立；当假定群体是异质性风险偏好时,在适度的参数条件下,个体风险偏好的异质性实现了合作策略与网络结构的共生演化,而且,随着风险偏好异质程度的增加,合作水平逐渐下降。同时,我们分别研究了迁徙成本、背叛诱惑参数和群体密度对合作水平的影响。最后,为了探讨基于社会福利的个体迁徙和合作演化之间的关系,我们应用计算经济学的方法研究了空间囚徒困境博弈中不同社会福利函数的迁徙模式对合作演化的影响。在文中,我们给出了三种福利函数：古典效用主义的社会福利函数(Utilitarian SWF)、贝尔努利-纳什社会福利函数(Bernoulli-Nash SWF)和罗尔斯社会福利函数(Rawlsian SWF)。仿真结果显示,在大部分参数空间内,基于社会福利的个体迁徙促使了合作行为的涌现。而且,在大部分的群体密度条件下,Utilitarian SWF和Rawlsian SWF下的合作水平要高于个体不迁徙的情况,但相比较个体不迁徙的情况,Bernoulli-Nash SWF抑制了合作行为的涌现。考虑社会福利是个体他涉偏好的一种体现,因此,我们的工作为理解社会系统中他涉偏好和合作之间的关系提供了有益的帮助,为社会偏好和合作演化之间的研究架起了桥梁,可以为政策制定者提供一定的参考依据。更多还原

【Abstract】 How social is possible, which is the basic issue of social science; how cooperation is possible, which is the fundamental question of behavioral science. Cooperation, which is widespread in social life, is the basis of human civilized society. However, why the altruism behavior can exist from an evolutionary perspective, is actually one unsolved mystery. Individuals’ migration in the social network plays an important role in the answering this question. The migration of agents is an important method for the formation of co-evolution and cooperative clumps. Meanwhile, the method of the computational economics provides a powerful tool for explaining the emergence of cooperation. On the basis of this context, this thesis uses computational method to study the influence of different migration mechanism on the evolution of cooperation.Evolutionary game theory as a theoretical tool and computational economics based agent as a research method are used to discuss the evolution of cooperation in the spatial Prisoner’s dilemma game from the view of the network structure and migration. To be more specific, we analysis the effect of the migration mechanism based on "to be close to the gentleman, to be away from the villain", heterogeneous risk preference and social welfare respectively, on the evolution of cooperation in the evolutionary game model. How the migration in the social network affects the cooperative behaviors, which has been explained in the following studies.First, we study the effect of the "to be close to the gentleman, to be away from the villain" on the evolution of cooperation. The meaning of the "to be close to the gentleman" is that agents cannot move when all of the neighbors are cooperators; otherwise, agents move with a probability which is the behavior for "to be away from the villain". we establish a migration model based on payoff and cost to discuss the effect of the migration cost on the cooperation level. The results show that the moving mechanism based on "to be close to the gentleman, to be away from the villain" improves the level of cooperation in a wide parameter space, which provides new ideas for understanding the cooperative behaviors in the society. The spatial state diagrams under different conditions have been shown to provide an intuitive understanding of the emergency for the cooperative clusters. Besides, we study the evolution of cooperation in the small world network in order to explore the effectiveness of the migration mechanism in other network structures. The results show that the migration mechanism promotes the cooperation in the small world network.Next, we study the effect of the migration based on risk preference on the evolution of cooperation in the spatial Prisoner’s dilemma game. Previous studies had not considered the risk preference or supposed the agents are risk neutral. However, the risk preference is heterogeneous in real life. So, we establish a migration model based on the risk preference and simulate the dynamic evolution process by using the computational economics. Moreover, we distinguish homogeneous risk preference influence on the cooperation from heterogeneous one. When we assume the homogeneity of risk preference, the results show that the risk aversion population is not beneficial for the emergence of cooperators; when we suppose the heterogeneity of risk preference; the results show that the heterogeneity of risk preference realizes the co-evolution of cooperative phenomenon and network structure under some parameters. Moreover, the level of cooperation reduces with the degrees of heterogeneity increasing. Meanwhile, we study the effect of the migration cost, the temptation to defect and population density on the cooperation level, respectively.Last, in order to study the relation between social welfare based migration and the evolution of cooperation, we consider an evolutionary Prisoner’s dilemma game(PDG) where an individual’s migration depends on different social welfare functions. We introduce three important social welfare functions (SWFs) commonly studied in social science:Utilitarian SWF, Bernoulli-Nash SWF and Rawlsian SWF. We find that social welfare based migration can promote cooperation under a wide range of parameter values. The cooperation level under Utilitarian SWF and Rawlsian SWF is higher than no migration, but Bernoulli-Nash SWF inhibits the emergency of cooperation compared with the no migration. The consideration of social welfare is a reflection of other-regarding of individuals, therefore, our work should be helpful to understand the evolution of human cooperation and bridge the chasm between the studies of social preferences and social cooperation. We can also provide some references for policy makers.更多还原