计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (16): 135-138.

• 数据库、信号与信息处理 • 上一篇    下一篇

一种基于MBR的高效的时间序列表示方法

孙梅玉1,2,唐 漾1,方建安1   

  1. 1.东华大学 信息科学与技术学院,上海 201620
    2.山东省工会管理干部学院 计算机系,济南 250100
  • 收稿日期:2007-09-07 修回日期:2007-11-26 出版日期:2008-06-01 发布日期:2008-06-01
  • 通讯作者: 孙梅玉

Improving indexing approach on time series based on minimum bounding rectangle

SUN Mei-yu1,2,TANG Yang1,FANG Jian-an1   

  1. 1.College of Information Science and Technology,Donghua University,Shanghai 201620,China
    2.Computer Department,Shandong Labour Union Administrators College,Jinan 250100,China
  • Received:2007-09-07 Revised:2007-11-26 Online:2008-06-01 Published:2008-06-01
  • Contact: SUN Mei-yu

摘要: 提出了一种基于最小边界矩形的新颖的时间序列表示方法(GMBR),该方法将网格的概念引入到MBR中,能够在保证低开小的情况下有效地提高查找的准确性,最后通过实验证明了该方法的有效性,实验分别在实际数据和合成数据上进行。结果表明该方法的剪枝率为69%~92%,高出MBR方法4%~9%。

关键词: GMBR, 表示方法, 时间序列, 数据挖掘, 相似性查找

Abstract: This paper proposess a novel time series representation called GMBR based on Minimum Bounding Rectangle in which the binary idea is applied into the MBR.The experiments have been performed on synthetic,as well as real data sequences to evaluate the proposed method.The experiment demonstrates that 69~92 percent of irrelevant sequences are pruned using the proposed method.

Key words: GMBR, representation, time series data, data mining, similarity search