中国科学技术大学学报 ›› 2014, Vol. 44 ›› Issue (12): 1024-1032.DOI: 10.3969/j.issn.0253-2778.2014.12.011

• 论著 • 上一篇    

群决策中决策发起人与专家之间的委托-代理均衡

高先务   

  1. 安徽建筑大学管理学院,安徽合肥 230022
  • 收稿日期:2013-06-21 修回日期:2013-06-23 接受日期:2013-06-23 出版日期:2013-06-23 发布日期:2013-06-23
  • 作者简介:高先务,男,1972年生,博士/讲师. 研究方向:决策科学与技术. E-mail:gxw968@sohu.com
  • 基金资助:
    国家自然科学基金重点项目(71231004)资助.

The equilibrium of principal-agent between principal and experts in group decision making

GAO Xianwu   

  1. School of Management, Anhui Jianzhu University, Hefei 230022, China
  • Received:2013-06-21 Revised:2013-06-23 Accepted:2013-06-23 Online:2013-06-23 Published:2013-06-23

摘要: 在决策发起人与专家利益不一致的群决策中,存在着两种博弈关系,一是决策发起人(委托人)与群专家(代理人)之间的委托-代理博弈;另一是专家之间的完全信息静态博弈.这里给出了3个基本公理,讨论了不同的专家净收益函数对委托人目标实现的影响,并选择了几种典型的博弈模型做对比分析.得出以下结论:委托人的最优选择是带偏差罚函数的博弈模型,且罚函数只对部分偏差做惩罚;专家的最优选择是持续投入,直至估值偏差满足委托人的阀值要求;最终,实现了群专家内部均衡和委托-代理均衡.委托-代理均衡满足激励相容约束,可以将决策发起人与专家的利益实现统一.

关键词: 群决策, 博弈, 投入-偏差函数, 委托-代理, 偏差罚函数

Abstract: There are two game relations in group-decision-making in which decision promoters and experts have inconsistent benefits, one is the relationship between promoters and experts, the other is static game of complete information relationship among experts. Three basic axioms were given, and the influence of a principals goal attainment on an experts payoff was discussed. Comparative analysis of several game models was carried out. The results show that, the principals optimal choice is the game model with experts payoff which contains a deviation penalty function. In addition, the domain of definition of deviation penalty function is only part of the deviation interval. The experts optimal choice is to invest constantly until the deviation of estimation satisfying the request of principal. Finally, the two equilibriums are both achieved. The principal-agent equilibrium satisfies incentive compatibility constraint, thus realizing the benefits of both the principal and the agent.

Key words: group decision making, game, input-deviation function, principal-agent, deviation penalty function

中图分类号: