节点文献
聚类分析系统的设计与实现及在工业中的应用
Design and Implement of Clustering Analysis System and Its Application in Industry
【作者】 宋辉;
【作者基本信息】 天津科技大学 , 检测技术及自动化装置, 2004, 硕士
【摘要】 本论文是国家863计划“冶金工业MES(制造执行系统)架构和关键技术研究与示范应用”的内容之一,所做的研究为数据挖掘的一个方法,具体内容是聚类分析系统的设计与实现及其在工业数据分析中的应用。针对工业数据中存在的数据量巨大、有噪声以及多模态性等特点,本论文提出了自组织映射神经网络和模糊C均值两种算法,从而为工业数据的分析提供了良好的数据挖掘的解决方案。 论文介绍了数据挖掘的相关知识、工业数据的特点以及数据挖掘技术在工业数据中应用的目的和意义,重点研究自组织映射神经网络和模糊C均值两种算法及其实现。 在WINDOWS平台上选用VC++6.0开发工具,并通过ADO接口与数据库进行连接,最终实现聚类分析系统,同时保证了系统的可视化和可操作性。运用实例对系统进行了分析和验证,结果表明本聚类系统对工业数据的分析达到了很好的效果。
【Abstract】 This paper belongs to one of the National 863 Plan, which researches and demonstrates the frame and the key technique of MES(Manufacturing Execution System) in metallurgy industry. This paper researches one method in data mining, which is Clustering Analyais System, designs and realizes the system, and uses it in analysing industry data. Because the industry data has the characters of largeness, noise and multi-mode, this paper puts forward SOFM(Self-Organizing Feature Map) and FCM(Fuzzy c-Means) to offer the data-analyse in industry a good data mining tool.This paper has the knowledge about data mining, the characters of industry data, the aim and signification of using data mining tool in industry. The most important research is the two arithmetics : SOFM and FCM and their realization. Programming the two arithmetics in VC++6.0 on WINDOWS, communicating database through ADO interface, this paper realizes the Clustering Analyais System, ensures its viaual and operational trait, this paper analyses and validate the system by examples,the results show that the system has a good effect in analysing industry data.
【Key words】 Data mining; Industry data; Clustering Analyais System; VC++6.0; SOFM(Self-Organizing Feature Map); FCM(Fuzzy c-Means);
- 【网络出版投稿人】 天津科技大学 【网络出版年期】2004年 04期
- 【分类号】TP399
- 【被引频次】5
- 【下载频次】332