Development of Decision Support Systems in Ambiguous Environments and Its Application

研究目的：本研究旨在提出在模糊环境中处理决策的方法。具体而言，本文将决策分析扩展到多目标决策和群体决策。研究计划：本文的研究内容和方法如下：1.构造概率多属性效用函数的决策支持系统的基础研究（1）基于经典决策分析及其模糊扩展，提出了可能性效用函数的构造方法。此外，作为概率效用模型的扩展，提出了可能性效用模型。在不完全信息结构中，该模型的使用也被用于评估风险前景的多重评价。(2)研究了建设性冲突解决的谈判分析，并沿着引入不确定性讨论了谈判区域的扩大.模糊数学规划是求解对策论问题最优解的一种方法，给出了处理模糊对策论问题各种模型的方法。讨论了两级决策系统中的交互式决策过程。不确定性决策分析的交互式计算机辅助决策支持系统的研制。(1)对本课题组开发的MAP和IDASS两个计算机程序进行了修改，并给出了新的讨论版本。(2)开发了一种新的计算机程序，该程序具有交互式评估和计算过程，可用于辅助不确定性多属性决策分析。

Aim of Research : This research concerns presenting the methodologies for treating decision making in ambiguous environments. In particular, We are concerned with extending decision analysis to multiobjective and group decisions.Executed Research Plan : The contents and the method of our research are as follows.1. Basic Research of decision support systems for constructing probabilistic multiattribute utility functions(1) Construction of possibilistic utility functions based on classical decision analysis and its fuzzy extensions are presented. In addition, the possibilistic utility models as an extension of the probabilistic utility models is presented. The use of this model in the incomplete information structure also is examined for assessing the multiple evaluation of risk prospects. (2) Negotiation analysis for constructive conflict solving is examined and the enlargement of the negotiation zone is discussed along with introducing uncertainty.2. Fuzzy mathematical programming approach for seeking the optimal solutions for game theoretic problems.The methods for treating various modeling for fuzzy game theoretic problems are presented. Interactive decision processes in the 2-level decision systems also are discussed.Development of interactive computer-aided decision support systems for decision analysis under uncertainty.(1) Existing computer programs, MAP and IDASS, which has been developed by our research group were revised and the new version of the discussions for those programs were presented. (2) A new computer program with interactive assessment and calculation processes for aiding the multiattribute decision analysis under uncertainty was developed.