流数据挖掘中数据流聚类算法研究  

The Study of Data Stream Clustering Algorithms in Data Stream Mining

作　　者：王高洋 李英梅[1] 

机构地区：[1]哈尔滨师范大学计算机科学与信息工程学院,哈尔滨150025

出　　处：《智能计算机与应用》2014年第5期6-8,13,共4页Intelligent Computer and Applications

基　　金：黑龙江省教育厅科学技术研究项目(12541239)

摘　　要：随着传感器数据、互联网数据、金融数据(股票价格等)、在线拍卖以及事务日志(网站访问日志、电话记录日志)等的不断产生,数据流成为了主要的数据形式。流挖掘是数据库领域的研究热点,有很大的应用前景。本文首先简单介绍了数据流与聚类分析的概念,阐述了数据流中的聚类分析及其要求,详细说明了主要传统聚类方法的演变及各自代表性流数据聚类算法,并对其进行总结。在本文的最后,对流数据挖掘的前景做出展望。With the producing of numerous data from financial tickers, network traffic monitoring, web and transaction log analysis, and sensor networks, the stream data has become a kind of main data type. Streaming mining is a hot issue in the career of database and it has a good future. In this article, firstly gives a brief introduction to data stream, clustering analysis and other concepts related to clustering data streaming. Secondly the paper elaborately states the main traditional cluste- ring methods and their typical clustering algorithms of data stream. After that, the paper also makes a summary to them. Finally, the paper makes the prospect of data mining.

关 键 词：聚类 数据流 数据流聚类 数据挖掘 数据流挖掘 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]