Gomez Portillo Ignacio
Departament de Física, Grup de Física Estadística, Universitat Autónoma de Barcelona, Barcelona 08193, Spain.
Phys Rev E Stat Nonlin Soft Matter Phys. 2012 Nov;86(5 Pt 1):051108. doi: 10.1103/PhysRevE.86.051108. Epub 2012 Nov 6.
We study the cooperation problem in the framework of evolutionary game theory by using the prisoner's dilemma as a metaphor of the problem. By considering the growing process of the system and individuals with imitation capacity, we show conditions that allow the formation of highly cooperative networks of any size and topology. By introducing general considerations of real systems, we reduce the required conditions for cooperation to evolve, which approaches the benefit-cost ratio r for the theoretical minimum r=1 when the mean connectivity of the individuals is increased. Throughout the paper, we distinguish different mechanisms that allow the system to maintain high levels of cooperation when the system grows by incorporation of defectors. These mechanisms require heterogeneity among individuals for cooperation to evolve. However, the required benefit-cost ratio and heterogeneities are drastically reduced as compared to those required for static networks.
我们以囚徒困境作为该问题的一种隐喻，在进化博弈论的框架下研究合作问题。通过考虑系统以及具有模仿能力的个体的增长过程，我们给出了允许形成任意规模和拓扑结构的高度合作网络的条件。通过引入对实际系统的一般考量，我们降低了合作得以演化所需的条件，当个体的平均连通性增加时，该条件趋近于理论最小值r = 1时的收益成本比r。在整篇论文中，我们区分了不同的机制，这些机制使得当系统通过纳入背叛者而增长时，系统能够维持高水平的合作。这些机制要求个体之间存在异质性才能使合作得以演化。然而，与静态网络所需的条件相比，所需的收益成本比和异质性大幅降低。
