主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

中国管理科学 ›› 2021, Vol. 29 ›› Issue (1): 24-35.doi: 10.16381/j.cnki.issn1003-207x.2021.01.003

• 论文 • 上一篇    下一篇

一类序贯决策方法的提出:透镜与直觉模糊信息的视角

顾婧1, 曾璐璇1, 李雯欣1, 徐泽水2   

  1. 1. 四川大学经济学院, 四川 成都 610065;
    2. 四川大学商学院, 四川 成都 610065
  • 收稿日期:2019-12-09 修回日期:2020-03-04 发布日期:2021-02-07
  • 通讯作者: 顾婧(1980-),女(汉族),江苏扬州人,四川大学经济学院,教授,博士生导师,博士,研究方向:风险投资决策、行为金融决策等,E-mail:gj0901@scu.edu.cn. E-mail:gj0901@scu.edu.cn
  • 基金资助:
    国家自然科学基金资助项目(71401116)

A Class of Sequential Decision Making Methods: Lens and Intuitionistic Fuzzy Information Perspective

GU Jing1, ZENG Lu-xuan1, LI Wen-xin1, XU Ze-shui2   

  1. 1. School of Economics, Sichuan University, Chengdu 610064, China;
    2. School of Business, Sichuan University, Chengdu 610064, China
  • Received:2019-12-09 Revised:2020-03-04 Published:2021-02-07

摘要: 与大多数关注经典秘书问题的序贯决策研究不同,本文针对决策对象不变而其状态更新的一类序贯决策问题,提供决策方法。相较已有研究而言,本文从透镜与直觉模糊信息的视角提出这类序贯决策的一般方法,更加具有普适性,可应用于应急管理、风险投资等多个方向。该决策方法具有如下特点:1)借鉴透镜模型,通过当期可观测的线索,预测这类序贯决策问题的下一期状态;2)考虑到决策环境具有不确定性,采用直觉模糊数对决策者获取的信息线索进行刻画;3)为最大化模拟现实决策者,在前景理论框架下提出状态更新的序贯决策方法。本文提出的决策方法依据在于决策者对决策对象的预测"前景"值和预测绩效,基于此提出了具体的决策步骤,并辅以案例进行分析。通过对不同决策情形下决策结果的比较分析和对案例的稳健性检验,验证了该决策方法的适用性和有效性。

关键词: 状态更新序贯, 决策透镜模型, 直觉模糊, 前景理论

Abstract: Different from most sequential decisionmaking studies that focus on classical secretary problems, a decisionmaking method is constructed in this paper for a class of sequential decisionmaking problems, in which the object is unchanged but its state keeps updating in every stage. The method we propose is based on lens model and intuitionistic fuzzy set theory, which in details are: 1) using lens model for reference, the state in the next stage is predicted through the proximal information in current stage; 2) considering the uncertainty of the decisionmaking environment, intuitionistic fuzzy numbers are used to describe the information cues obtained by the decision-maker; 3) to better simulate the real decision-maker, the method is proposed in the framework of prospect theory. Based on above idea, two key parameters are invented for the method, named predicted "prospect" value and prediction performance, and specific decisionmaking steps are proposed. At last, a case in venture capital field is applied to illustrate the performance of the proposed method, where the data comes from analog simulation. Furthermore, comparative analysis of different occasion decision results and the robustness test are taken to verify the applicability and effectiveness of our method. Compared with the existing studies, the method we propose is more universal and can be applied to many scenes such as venture capital decisionmaking, emergency management and so on. Also, more research attention may be gathered in this field due to our study.

Key words: state update, sequential decision, lens model, intuitionistic fuzzy, prospect theory

中图分类号: