Enhanced Weighted Dynamic Time Warping for Time Series Classification

  • Pichamon AnantasechEmail author
  • Chotirat Ann Ratanamahatana
Conference paper
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 797)


Dynamic time warping (DTW) has been widely used as a distance measure for time series classification because its matching is elastic and robust in most cases. However, DTW may lead to over compression that could align too many consecutive points from one time series to only one point on another. As a result, important feature information could be overlooked, which can be the cause of misclassification particularly when the shape of time series is an essential feature. In order to fix this problem and improve the classification accuracy, we propose a distance measure called an enhanced weighted dynamic time warping, where weight functions are proposed and applied to the DTW distance measure. Other than being parameter-free, our experiment results have demonstrated to impressively outperform other rival methods by a large margin while having less time complexity than the state-of-the-art approaches.


Dynamic time warping Weighted dynamic time warping Adaptive weight Time series classification 


  1. 1.
    Berndt DJ, Clifford J (1994) Using dynamic time warping to find patterns in time series. In: Proceedings of the workshop on knowledge discovery in databases, pp 359–370Google Scholar
  2. 2.
    Legrand B, Chang CS, Ong SH et al (2008) Chromosome classification using dynamic time warping. Pattern Recognit Lett 29:215–222CrossRefGoogle Scholar
  3. 3.
    Rath TM, Manmatha R (2003) Word image matching using dynamic time warping, In: 2003 IEEE Computer society conference on computer vision and pattern recognition proceedings, vol 2, pp 521–527, IEEEGoogle Scholar
  4. 4.
    Bashir M, Kempf J (2008) Reduced dynamic time warping for handwriting recognition based on multidimensional time series of a novel pen device. Int J Intell Syst Technol (WASET) 3:1839–1845Google Scholar
  5. 5.
    Agrawal R, Faloutsos C, Swami A (1993) Efficient similarity search in sequence databases, In: Proceeding of Intl Conference on Data Organazation. 730, pp 69–84Google Scholar
  6. 6.
    Struzik ZR, Siebes A (1998) Wavelet transform in similarity paradigm, In: Proceedings of the second Pacific-Asia conference on research and development in knowledge discovery and data mining, pp 295–309CrossRefGoogle Scholar
  7. 7.
    Berndt DJ, Clifford J (1994) Using dynamic time warping to find patterns, In: Time series, KDD workshop. 10, pp 359–370Google Scholar
  8. 8.
    Berndt DJ, Clifford J (1996) Finding patterns in time series: a dynamic programming approach, In: Advances in knowledge discovery and data mining, American association for artificial intelligence, pp 229–248Google Scholar
  9. 9.
    Keogh EJ, Pazzani MJ (2001) Derivative dynamic time warping. In: Proceedings of the 2001 SIAM international conference on data miningGoogle Scholar
  10. 10.
    Yu D, Yu X, Hu Q et al (2011) Dynamic time warping constraint learning for large margin nearest neighbor classification. Inform Sci 181:2787–2796CrossRefGoogle Scholar
  11. 11.
    Jeong YS, Jeong MK, Omitaomu OA (2011) Weighted dynamic time warping for time series classification. Pattern Recogn 44(9):2231–2240CrossRefGoogle Scholar
  12. 12.
    Wan Y, Chen XL, Shi Y (2017) Adaptive cost dynamic time warping distance in time series analysis for classification. J Comput Appl Math 319:514–520MathSciNetCrossRefGoogle Scholar
  13. 13.
    Chen Y, Keogh E, Hu B, Begum N, Bagnall A (2015) Abdullah Mueen and Gustavo Batista (2015) The UCR Time Series Classification Archive. URL

Copyright information

© Springer Nature Singapore Pte Ltd. 2019

Authors and Affiliations

  • Pichamon Anantasech
    • 1
    Email author
  • Chotirat Ann Ratanamahatana
    • 1
  1. 1.Department of Computer EngineeringChulalongkorn UniversityBangkokThailand

Personalised recommendations