群决策下模糊决策粗糙集的冲突分析及其应用方法研究

The decision problem with fuzzy uncertainty and its conflict is becoming the centre of people’s attention. One of its core content is to study how to use scientific methods to accommodate the real decision-making needs. As a novel method, decision-theoretic rough sets (DTRS) not only have the advantages in dealing with uncertainty decision problems, but also can measure decision risks using loss function. Nowadays, DTRS have been successfully applied to many domains. This project focuses on the model of DTRS and selects its loss function as the breakthrough point. By introducing the fuzzy numbers into DTRS, we systematically study the consistency and application methods of fuzzy decision-theoretic rough sets (FDTRS) in the group decision-making. In light of this framework, the main contents of this project are organized as follows: 1. The model of FDTRS is constructed with the aid of group decision-making. Meanwhile, we solve the inconsistency problem of group decision-making and analyze the ranking methods of fuzzy numbers. These works are the basis of conflict analysis and the applications. 2. The conflict rules derived from FDTRS under the two-categories and multi-categories problems are analyzed, respectively. The coordination mechanisms of conflict rules are designed, which ensure the consistency of decision rules. 3. Considering the risk factors, we further explore the application methods of FDTRS based on decision rules, e.g. ranking, resource allocation and attribute reduction. This project can make DTRS adapt to the fuzzy environment and solve conflict problems presented in DTRS, which pushes the application researches of DTRS.

决策问题的模糊不确定性及其所带来的冲突，是当前人们普遍关注的问题，其核心内容之一是研究如何利用科学方法来满足现实需求。决策粗糙集作为一种新的处理不确定性决策问题的方法，由损失函数测度决策风险，已成功应用到多个管理领域。本项目以决策粗糙集为研究对象，由损失函数为研究切入点，将模糊数引入决策粗糙集中，系统研究群决策下模糊决策粗糙集的冲突分析及其应用方法。主要内容包括：1. 建立模糊环境下基于群决策的模糊决策粗糙集模型，解决群决策的不一致性并讨论模糊数的排序方法，为分析冲突规则和应用奠定基础；2. 研究两分类和多分类决策问题下模糊决策粗糙集的冲突规则，分析冲突机理并设计协调机制，使之在实际应用中保持决策依据的一致性；3. 考虑风险代价，进一步探索基于决策规则的排序、资源分配和属性约简等应用方法。本项目不仅使决策粗糙集适应模糊环境，还解决该模型中存在的冲突问题，具有重要的应用价值。

决策粗糙集作为一种新的处理不确定性决策问题的方法，由损失函数测度决策风险，已成功应用到多个管理领域。决策问题的模糊不确定性及其所带来的冲突，是当前人们普遍关注的问题，其核心内容之一是研究如何利用科学方法来满足现实需求。本项目以决策粗糙集为研究对象，由损失函数为研究切入点，将模糊数引入决策粗糙集中，系统研究群决策下模糊决策粗糙集的冲突分析及其应用方法。. 通过三年来的研究，项目组已在国内外重要期刊上发表(录用)论文近20篇，其中3篇入选ESI热点论文，2篇入选ESI高被引用论文，SCI检索15篇，SSCI检索1篇，参与编著2部。本项目主要研究内容有：1. 建立模糊环境下基于群决策的模糊决策粗糙集模型，解决群决策的不一致性并讨论模糊数的排序方法，为分析冲突规则和应用奠定基础；2. 研究两分类和多分类决策问题下模糊决策粗糙集的冲突规则，分析冲突机理并设计协调机制，使之在实际应用中保持决策依据的一致性；3. 考虑风险代价，进一步探索基于决策规则的风险评价与排序等应用问题。此外，本项目的相关应用成果获得了四川省2015年科技进步二等奖和四川省第十七次哲学社会科学优秀成果三等奖各1项，项目已得到高质量地完成。. 本项目不仅使决策粗糙集适应模糊环境，还解决该模型中存在的冲突问题，具有重要的应用价值。