Co-evolution of payoff strategy and interaction strategy in prisoner's dilemma game
Co-evolution of payoff strategy and interaction strategy in prisoner's dilemma game
复制标题
囚徒困境博弈中支付策略与交互策略的协同演化
DOI：
10.1016/j.physa.2016.04.045
复制
发表时间：
2016
影响因子：
3.3
通讯作者：
Cheng Hongyan
中科院分区：
文献类型：
--
作者：
Zhang Kangjie;Cheng Hongyan
协同进化动力学模型为研究复杂系统提供了一个现实的范式，得到了广泛的研究。研究了支付策略和交互策略的协同演化问题。从收益策略r的初始高斯分布开始，均值为u，方差为q，我们关注收益策略的最终分布。我们发现，最终分布的回报策略可能会显示不同的结构，这取决于参数。在u<− 1和u> 3的范围内，分布显示出关于r= u对称的单峰结构。该分布在− 1< u< 3的范围内表现为双峰结构，尽管在1< u< 2的范围内出现了假的三峰结构。对不同类型的收益策略分布的形成给出了解释。
Co-evolutionary dynamical models, providing a realistic paradigm for investigating complex system, have been extensively studied. In this paper, the co-evolution of payoff strategy and interaction strategy is studied. Starting with an initial Gaussian distribution of payoff strategy r with the mean u and the variance q, we focus on the final distribution of the payoff strategy. We find that final distribution of the payoff strategy may display different structures depending on parameters. In the ranges u<− 1 and u> 3, the distribution displays a single-peak structure which is symmetric about r= u. The distribution manifests itself as a double-peak structure in the range− 1< u< 3 although a fake three-peak structure shows up in range 1< u< 2. The explanations on the formation of different types of payoff strategy distributions are presented.