计算机科学 ›› 2016, Vol. 43 ›› Issue (5): 234-237.doi: 10.11896/j.issn.1002-137X.2016.05.043

• 人工智能 • 上一篇    下一篇

一种基于关键点的时间序列线性表示方法

陈帅飞,吕鑫,戚荣志,王龙宝,余霖   

  1. 河海大学计算机与信息学院 南京211100,河海大学计算机与信息学院 南京211100,河海大学计算机与信息学院 南京211100,河海大学计算机与信息学院 南京211100,河海大学计算机与信息学院 南京211100
  • 出版日期:2018-12-01 发布日期:2018-12-01
  • 基金资助:
    本文受国家自然科学基金面上项目(61272543),国家科技支撑计划(2013BAB06B04),国家自然科学基金委-广东联合项目(U1301252),江苏省博士后科研资助

Linear Representation Method Based on Key Points for Time Series

CHEN Shuai-fei, LV Xin, QI Rong-zhi, WANG Long-bao and YU Lin   

  • Online:2018-12-01 Published:2018-12-01

摘要: 时间序列数据具有规模大、维度高等特点,直接在原始序列上进行数据挖掘,其计算复杂度高且易受噪声影响,因此对原始时间序列进行预处理是必不可少的,而常用的线性表示方法大多存在对分段点的筛选准确度不高的问题。基于时间序列的变化特征,提出了一种基于时间序列关键点的线性表示方法。该方法综合考虑了时间跨度和振幅变化,能高效提取时间序列中的关键点,并防止过度除噪,实现简单。实验表明,该方法对不同领域的数据具有良好的普适性。

关键词: 数据挖掘,时间序列,线性表示,关键点,过度除噪

Abstract: Time series data has the features of large scale and high latitude.It has high computational complexity and is susceptible to noise if doing data mining on the raw sequence directly,so the original time series pretreatment is essential,and most methods of commonly used linear representation have low accuracy in selection piecewise points.Based on the time series variation,we proposed a linear representation method based on key points for time series.The method takes into account the time span and amplitude changes and can efficiently extract key points in the time series,which can prevent excessive noise removal and is implemened simply.Experiments show that the method has good universality for data from different areas.

Key words: Data mining,Time series,Linear representation,Key points,Excessive noise removal

[1] Pan Ding,Shen Jun-yi.Similarity Discovery Techniques in Temporal Data Mining[J].Journal of Software,2007,18(2):246-258(in Chinese) 潘定,沈钧毅.时态数据挖掘的相似性发现技术[J].软件学报,2007,18(2):246-258
[2] Keogh E.Fast similarity search in the presence of longitudinal scaling in time series databases[C]∥Proceedings of the International Conference on Tools with Artificial Intelligence,1997.Washington:IEEE Computer Society,1997:578-584
[3] Das G,Lin K I,Mannila H,et al.Rule Discovery from Time Series[C]∥KDD-98.New York:KDD,1998:16-22
[4] Debrégeas A,Hébrail G.Interactive Interpretation of Kohonen Maps Applied to Curves[C]∥KDD-98.New York:KDD,1998:179-183
[5] Hellerstein J M,Koutsoupias E,Papadimitriou C H.On the analysis of indexing schemes[C]∥Proceedings of the ACM SIGACT-SIGMOD-SIGART Symposium on Principles of Database Systems-PODS,1997.Tucson:ACM,1997:249-256
[6] Agrawal R,Faloutsos C,Swami A.Efficient similarity search in sequence databases[C]∥Proceedings of the 4th International Conference on Foundations of Data Organization and Algorithms,1993.London:Springer Berlin Heidelberg,1993:69-84
[7] Chan K P,Fu A W C.Efficient time series matching by wavelets[C]∥Proceedings International Conference on Data Enginee-ring,1999.Sydney:IEEE,1999:126-133
[8] Keogh E,Chakrabarti K,Pazzani M,et al.Dimensionality reduction for fast similarity search in large time series databases[J].Knowledge and information Systems,2001,3(3):263-286
[9] Yi B K,Faloutsos C.Fast time sequence indexing for arbitrary Lp norms[C]∥Proceedings of the 26th VLDB Conference,2000.Cairo:VLDB,2000:385-394
[10] Wu D,Singh A,Agrawal D,et al.Efficient retrieval for browsing large image databases[C]∥International Conference on Information and Knowledge Management,1996.Rockville:ACM,1996:11-18
[11] Pratt K B,Fink E.Search for patterns in compressed time series[J].International Journal of Image and Graphics,2002,2(1):89-106
[12] Xiao H,Feng X F,Hu Y F.A new segmented time warping distance for data mining in time series database[C]∥Proceedings of 2004 International Conference on Machine Learning and Cybernetics,2004.Shanghai:IEEE,2004:1277-1281
[13] Liu Shi-yuan,Jiang Hao.A Review on Time Series Representation for Similarity-based Pattern Search[J].Computer Engineering and Applications,2004,0(27):53-59(in Chinese) 刘世元,江浩.面向相似性搜索的时间序列表示方法述评[J].计算机工程与应用,2004,40(27):53-59
[14] Zhan Yan-yan,Xu Rong-cong,Chen Xiao-yun.Time Series Pie-cewise Linear Representation Based on Slope Extract Edge Point[J].Computer Science,2006,3(11):139-142(in Chinese) 詹艳艳,徐荣聪,陈晓云.基于斜率提取边缘点的时间序列分段线性表示方法[J].计算机科学,2006,33(11):139-142
[15] Perng C S,Wang H,Zhang S R,et al.Landmarks:a new model for similarity-based pattern querying in time series databases[C]∥Proceedings International Conference on Data Engineering,2000.San Diego:IEEE,2000:33-42

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!