基于差别矩阵和重要度的增量式属性约简算法  

Incremental Algorithm for Attribute Reduction Based on Differential Matrix and Importance

作　　者：高晓红 李兴奇 GAO Xiaohong;LI Xingqi(School of Mathematics and Statistics,Chuxiong Normal University,Chuxiong 675000,China;School of Economics and Management,Chuxiong Normal University,Chuxiong 675000,China)

机构地区：[1]楚雄师范学院数学与统计学院,云南楚雄675000 [2]楚雄师范学院经济与管理学院,云南楚雄675000

出　　处：《长春大学学报》2020年第6期15-23,共9页Journal of Changchun University

基　　金：国家自然科学基金资助项目(11261001);云南省应用基础研究计划青年项目(2017FD152)。

摘　　要：已有的基于差别矩阵的属性约简算法时空复杂度高,并且大多数主要是针对决策表(或信息系统)不变的情况,关于属性约简的增量式更新算法研究还不多。因此,提出了一种基于差别矩阵和属性重要度的增量式属性约简算法,主要解决条件属性增加情况下的属性约简求解问题。该算法结合差别矩阵和属性依赖度,从属性依赖度的角度出发度量了属性重要度,最终求得属性约简集。理论分析及实验结果表明,所提算法是有效可行的,提高了属性约简效率,明显降低了时间和空间复杂度。The complexity of time and space of the existed algorithms for attribute reduction based on differential matrix is high,and many algorithms mainly aim at the case of stationary decision table or information system,very little work has been done in the updating of algorithms for attribute reduction.Therefore,an incremental algorithm for attribute reduction based on differential matrix and importance is proposed,which is mainly used to solve attribute reduction when condition attributes are dynamically increased.Differential matrix and attribute dependence are used in the new algorithm,and the attribute importance is measured from the angle of attribute dependence,finally,an attribute reduction set is obtained.Theoretical analysis and experimental results have shown that this new algorithm is effective and feasible,having improved the efficiency of attribute reduction and reduced the time and space complexity significantly.

关 键 词：粗糙集 差别矩阵 属性重要度 属性约简 增量式算法 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]