论文信息 - Regression time warping for similarity measure of sequence

Regression time warping for similarity measure of sequence

In the paper, we propose regression time warping (RTW), a similarity measure for sequence or time series matching. RTW fuses the linear regression analysis, which controls the shifting and scaling factors between sequences (Lei and Govindaraju, 2004), and the principles of dynamic time warping (DTW), which provides robustness with elastic matching. RTW has complexity as low as O(n), while the complexity of DTW is O(n/sup 2/). Experimental results show the accuracy of RTW in classification is comparable to DTW, and much faster than DTW in term of running time.

Venu Govindaraju | Hansheng Lei

[1] V. Govindaraju,et al. GRM: Generalized regression model for clustering linear sequences , 2004 .

[2] Eamonn J. Keogh,et al. Scaling up dynamic time warping for datamining applications , 2000, KDD '00.

[3] Jeffrey M. Wooldridge,et al. Introductory Econometrics: A Modern Approach , 1999 .

[4] Wesley W. Chu,et al. An index-based approach for similarity search supporting time warping in large sequence databases , 2001, Proceedings 17th International Conference on Data Engineering.

[5] Donald J. Berndt,et al. Using Dynamic Time Warping to Find Patterns in Time Series , 1994, KDD Workshop.

[6] Eamonn J. Keogh,et al. UCR Time Series Data Mining Archive , 1983 .

[7] Eamonn Keogh. Exact Indexing of Dynamic Time Warping , 2002, VLDB.

[8] Christos Faloutsos,et al. Efficient retrieval of similar time sequences under time warping , 1998, Proceedings 14th International Conference on Data Engineering.