集值决策信息系统的增量式属性约简算法  

Incremental Attribute Reduction Algorithm of the Set-valued Decision Information System

作　　者：王映龙[1,2] 华佳佳 钱文彬 柳军[2] WANG Ying-long;HUA Jia-jia;QIAN Wen-bin;LIU Jun(School of Computer and Information Engineering, Jiangxi Agricultural University, Nanchang 330045, China;School of Software, Jiangxi Agricultural University, Nanchang 330045, China;Key Laboratory of Agricultural Information Technology of Jiangxi Province, Nanchang 330045, China)

机构地区：[1]江西农业大学计算机与信息工程学院,南昌330045 [2]江西省高等学校农业信息技术重点实验室,南昌330045 [3]江西农业大学软件学院,南昌330045

出　　处：《小型微型计算机系统》2018年第6期1239-1244,共6页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61502213;71461013;61462038)资助;江西省自然科(20151BAB217009;20132BAB201045)资助;江西省教育厅科学技术项目(GJJ150399)资助

摘　　要：在现实应用中许多数据往往呈动态变化,非增量式约简方法处理此类数据需消耗大量的计算时间和存储空间.增量式技术能有效处理动态变化的数据.针对集值决策信息系统中数据的动态变化情况,研究了新增对象对信息系统中原有知识的影响,分析了新增对象对分布约简的更新机制,提出了一种增量求解分布函数的计算方法,在此基础上,设计了集值决策信息系统的增量式属性约简算法,当新增对象加入到决策信息系统时,算法能够利用原决策信息系统的约简知识,快速更新分布协调集,通过计算极小析取范式得到属性约简;最后,通过实例验证分析了算法的可行性和有效性.In practical applications,many data often change dynamically. Processing these data,non-incremental reduction method will cost lots of computing time and storage space. However,incremental technique can effectively handle these data. For the dynamic change of data in set-valued decision information system,we study that new object has impact on prior knowledge in the information system. And the updating mechanism of new object to the distributed reduction is analyzed. Then,a computational method for incremental distribution function is proposed. On this basis,incremental attribute reduction algorithm of the set-valued decision information system is designed. When the new object is added to the decision information system,the algorithm can update the distributed coordination set quickly by using the reduction knowledge of the original decision information system and achieve attribute reduction by computing minimal disjunctive normal form. Finally,the effectiveness and feasibility of the proposed algorithm are validated by the examples.

关 键 词：粗糙集 属性约简 增量学习 分布协调集 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]