基于可分离性的多尺度决策信息系统的最优尺度约简  

Optimal Scale Reduction in Multi-scale Decision Information Systems Based on Separability

作　　者：金铭 陈锦坤 李进金[1,2] JIN Ming;CHEN Jinkun;LI Jinjin(School of Mathematics and Statistics,Minnan Normal University,Zhangzhou 363000,China;Fujian Key Laboratory of Granular Computing and Applications,Minnan Normal University,Zhangzhou 363000,China)

机构地区：[1]闽南师范大学数学与统计学院,福建漳州363000 [2]闽南师范大学福建省粒计算及其应用重点实验室,福建漳州363000

出　　处：《郑州大学学报（理学版）》2024年第4期72-80,共9页Journal of Zhengzhou University:Natural Science Edition

基　　金：国家自然科学基金项目(62076116,11871259);福建省自然科学基金项目(2020J01792,2021J02049,2020J02043)。

摘　　要：针对最优尺度约简问题,从对象与决策类的关系出发,提出一种基于可分离性的多尺度决策信息系统的约简方法。首先,分别给出类内对象紧性和类间对象分散度的定义并探究其性质。其次,在多尺度决策信息系统中通过类内对象紧性和类间对象分散度定义属性子集的可分离性,并给出可分离性与约简之间的关系,在此基础上,结合属性权重与尺度权重给出了基于可分离性的重要度。最后,设计了一种基于重要度的启发式最优尺度约简算法。实验结果表明,所提方法在分类精度和约简集基数上具有较大的优势。Aiming at the problem of optimal scale reduction,a reduction method in multi-scale decision information systems based on separability was proposed from the relationship between objects and decision classes.Firstly,the degree of aggregation of intraclass objects and the degree of dispersion of between-class objects were defined and their properties were explored.Secondly,the separability of attribute subset was defined by the degree of aggregation of intraclass objects and the degree of dispersion of between-class objects in multi-scale decision information systems,and the relationship between separability and reduction was given.On this basis,the importance degree based on separability was given by combining attribute and scale weights.Finally,a heuristic optimal scale reduction algorithm based on importance degree was designed.The results showed that the proposed method had great advantages in classification accuracy and the cardinality of reduction set.

关 键 词：多尺度决策信息系统 最优尺度约简 可分离性 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]