基于可辨识矩阵的属性约简算法及应用  

Attribute Reduction Algorithm Based on Discernibility Matrix and Its Application

作　　者：陈志恩[1] 田彦山[1] 马旭[1] CHEN Zhi-en;TIAN Yan-shan;MA Xu(Department of Mathematics and Computer Science,Ningxia Normal College,Guyuan Ningxia 756000,China)

机构地区：[1]宁夏师范学院数学与计算机科学系,宁夏固原756000

出　　处：《大学数学》2021年第3期20-24,共5页College Mathematics

基　　金：宁夏高等学校科学技术研究第一资助项目(NGY2020083);宁夏高等学校一流学科建设(教育学学科资助项目)(NXYLXK2017B11);宁夏重点研发计划项目(2018BEE03025)。

摘　　要：针对决策信息系统最大分布约简问题,从代数角度给出了一种启发式属性约简算法.该算法在最大分布可辨识属性矩阵基础上,首先以最大分布核属性集为起点,然后对其余属性按其在可辨识属性矩阵中出现的频数大小逐次添加到核属性集中,再根据启发式算子对新的属性集给出最大分布约简的判断.重复以上步骤,直到找到最大分布约简.算例分析表明该算法的有效性和可行性.Aiming at the problem of maximum distribution reduction of decision information system,this paper proposes a heuristic attribute reduction algorithm from the algebraic point of view.Based on the distribution of the maximum discernibility attribute matrix,the algorithm firstly takes the maximum distributed core attribute set as the starting point.The remaining attributes are then successive added to the core attribute set according to the frequency of their occurrence in the identifiable attribute matrix.Thirdly,the maximum distribution reduction of the new attribute set is determined according to the attribute reduction algorithm.The steps above are repeated until find the maximum distribution reduction.The algorithm is proved validity and feasibility by an example.

关 键 词：信息系统 可辨矩阵 最大分布约简 

分 类 号：O23[理学—运筹学与控制论]