优势关系下带属性权重模糊决策表的部分一致约简  

Partially Consistent Reduction in Fuzzy Decision Table with Attribute Weights Based on Dominance Relation

作　　者：冉钦文 徐伟华 RAN Qin-wen;XU Wei-hua(College of Artificial Intelligent,Southwest University,Chongqing 400715,China)

机构地区：[1]西南大学人工智能学院,重庆400715

出　　处：《模糊系统与数学》2021年第6期28-35,共8页Fuzzy Systems and Mathematics

基　　金：国家自然科学基金资助项目(61976245,61772002)。

摘　　要：在现实生活中很多信息表是基于优势关系的,同时信息表中每个样本的属性值是模糊数,并且对不同的属性的重视程度也是不完全相同的,即对于不一样的属性存在着不一样的权重。针对这一情况,本文借鉴了直觉模糊序表的权重函数的想法,建立了一种在条件属性集上的权重函数,并给出了不协调的带属性权重的模糊序决策表。然后,在该决策表中定义部分一致函数和部分一致协调集,并利用部分一致可辨识矩阵的方法得到部分一致约简。最后,通过案例分析该方法的有效性。In real life,there are many information tables based on the dominance relationship,and the attribute value of each sample is fuzzy number,and the importance of different attributes is not the same.That is,different attributes have different weights.In view of this situation,this paper uses the idea of weight function of intuitionistic fuzzy order list for reference,establishes a weight function on conditional attribute set,and gives incongruous fuzzy order decision table with attribute weight.Then,the partially consistent function and partially consistent coordination set are defined in the decision table,and the partially consistent reduction is obtained by using the method of partially consistent identifiable matrix.Finally,a case is given to analyze the effectiveness of the method.

关 键 词：部分一致约简 权重函数 属性权重 决策表 

分 类 号：C931[经济管理—管理学]