摘要
协调博弈作为一类特殊的博弈,在过去几年受到广泛关注,研究主要集中在协调失败和适应性行为两方面。针对Huyck等研究个体适应性行为的局限性,将近视理论和虚拟动态理论进行扩展。基于实验调查数据,分别找出适合低均衡组和高均衡组的自适应行为模型,同时研究单个个体的适应性行为。研究发现,对于低均衡组,虚拟动态理论是预测中位数最适合的模型,近视理论是预测个体行动最适合的理论模型;对于高均衡组,虚拟动态理论在预测中位数和个体行动方面命中率都是最低的,且大部分受试者基于近视理论预测行动的命中次数最大。该研究丰富了适应性行为研究的相关理论,同时表明了个人利益和团队利益一致的重要性,慎重对待每一次选择并尽可能多参考有价值的信息,才能实现收益最大化。
As a special kind of game, the coordination game has received wide attention in the past few years. At present, research on the coordination game mainly focuses on coordination failure and adaptive behavior. Aiming at the limitations in the research of Huyck et al. , this paper extends myopic best response and fictitious play dynamic. To find out the adaptive behavior models for low equilibrium group and high equilibrium group, based on experimental data, different models are used to predict median and individual action in two groups and the accuracy of predicted results are compared. Meanwhile, individual adaptive behavior is studied. Results show that as to the low equilibrium group, fictitious play dynamic is the most suitable model for forecasting the median, and myopic best response is the most suitable theoretical model for predicting individnal action. As to the high equilibrium group, fictitious play dynamic is the lowest in the forecast of the median and individual action and most of the subjects based on myopic best response for predicting actions make the maximum hit rate. This study enriches the theories of adaptive behavior and demonstrates the importance of unanimity between personal interests and team interests. Everyone should treat every choice carefully and refer to valuable information as much as possible, therefore the benefits can be maximized.
作者
石文华
张绮
绳娜
SHI Wen-hua ZHANG Qi SHENG Na(School of Economics and Management, Beijing University of Posts and Telecommunications, Beijing 100876, China)
出处
《北京邮电大学学报（社会科学版）》
2017年第3期76-84,共9页
Journal of Beijing University of Posts and Telecommunications(Social Sciences Edition)
关键词
分水岭博弈
近视理论
虚拟动态理论
适应性行为
median game
myopic best response
fictitious play dynamic
adaptive behavior
