融合粗糙集与模糊聚类的连续数据知识发现.pdfVIP

第24 卷 第6 期 中 国 电 机 工 程 学 报 Vol.24 No.6 Jun. 2004 2004 年6 月 Proceedings of the CSEE ©2004 Chin.Soc.for Elec.Eng. 0258-8013 (2004) 06-0205-06 TK38 A 470 2099 文章编号： 中图分类号： 文献标识码： 学科分类号： ⋅ 融合粗糙集和模糊聚类的连续数据知识发现 于达仁，胡清华，鲍 文 （哈尔滨工业大学热能动力工程系，黑龙江哈尔滨 150001 ） COMBINING ROUGH SET METHODOLOGY AND FUZZY CLUSTERING FOR KNOWLEDGE DISCOVERY FROM QUANTITATIVE DATA YU Da-ren, HU Qing-hua ，BAO Wen (Harbin Institute of Technology, Harbin 150001, China) ABSTRACT: Knowledge acquisition is the technique 优化运行模式与诊断知识，是解决基于知识的决策 bottleneck of application of knowledge-based systems. Rough 与诊断系统知识获取瓶颈的有效方法。一般而言， set is an emerging tool to extract knowledge from a vast volume 电力系统的历史数据可以分为两种类型，一种是连 of data. Knowledge acquisition technique based on rough set is 续取值的定量变量，如电流、压力、温度等；另一 just suitable for discrete data, but most of raw data in power plants are continuous. To deal with the problem fuzzy 种是离散变量，这种属性的值域是有限个模糊语言 clustering is introduced to discretize the continuous attributes. 或者布尔代数，如“好、坏”，“高、中、低”，“开、 Fuzzy c-mean clustering algorithm is used in our work. A 关”等。对于连续数据，人们往往采用系统辨识、 clustering center vector and a membership matrix can be gotten 时间序列分析或者别的函数拟合的方法来学习隐 with the algorithm. By this way the continuous attributes is 含在数据中的规律。对于离散变量，机器学习的研 discretizated into decision table, which reduction and rule mining technique based on rough set theory can deal with. 究开发了一系列算法来挖掘数据之间的依赖关系， Some measures, support and confidence, evaluating the rough 如AQ11 ，ID3 等基于符号的归纳式机器学习方法。 rules are analyz