区间值目标信息系统属性约简

发布时间：2018-01-16 16:22

本文关键词：区间值目标信息系统属性约简　出处：《闽南师范大学》2017年硕士论文　论文类型：学位论文

【摘要】：粗糙集理论和模糊集理论都是用来解决不确定性、模糊性问题的两种重要的数学方法。本文应用这两种理论,在优势-等价关系上研究了区间值目标信息系统属性约简问题;在变精度优势关系上研究了区间值模糊目标信息系统的规则提取问题。主要工作如下:1.在优势-等价关系的区间值目标信息系统上,描述了分配约简、近似约简以及分布约简、最大分布约简的概念,探讨了它们各自之间的关系,并给出了它们之间的判定定理和辨识矩阵,从而得到在优势-等价关系下的区间值目标信息系统的属性约简的计算方法。2.在变精度优势关系下区间值模糊目标信息系统中,考虑了此关系下的属性约简问题,并在规则提取方面上,利用优势粗糙集隶属函数,得到每一个约简属性在不同决策类中的可能值,从而获得所需要的决策规则,使提取的决策更简便有效,更具有客观性。
[Abstract]:Rough set theory and fuzzy set theory are two important mathematical methods to solve the problem of uncertainty and fuzziness. The attribute reduction problem of interval valued object information system is studied on the advantage-equivalence relation. The rule extraction problem of interval valued fuzzy target information system is studied in variable precision dominance relation. The main work is as follows: 1. On the interval valued objective information system with advantage equivalent relation, assignment reduction is described. The concepts of approximate reduction, distribution reduction and maximum distribution reduction are discussed. The relationship between them is discussed, and the judgment theorem and identification matrix between them are given. The method of attribute reduction for interval valued object information system under the advantage-equivalence relationship is obtained. 2. In the interval valued fuzzy object information system with variable precision dominance. The problem of attribute reduction in this relation is considered. In the aspect of rule extraction, the possible values of each reduction attribute in different decision classes are obtained by using the membership function of superior rough set. In order to obtain the necessary decision-making rules, the extracted decision is more simple and effective, more objective.
【学位授予单位】：闽南师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;O159

【参考文献】