TR2017-168

Matrix Profile VII: Time Series Chains: A New Primitive for Time Series Data Mining


    •  Zhu, Y., Imamura, M., Nikovski, D.N., Keogh, E., "Matrix Profile VII: Time Series Chains: A New Primitive for Time Series Data Mining", International Conference on Data Mining, DOI: 10.1109/ICDM.2017.79, November 2017.
      BibTeX TR2017-168 PDF
      • @inproceedings{Zhu2017nov,
      • author = {Zhu, Yan and Imamura, Makoto and Nikovski, Daniel N. and Keogh, Eamonn},
      • title = {Matrix Profile VII: Time Series Chains: A New Primitive for Time Series Data Mining},
      • booktitle = {International Conference on Data Mining},
      • year = 2017,
      • month = nov,
      • doi = {10.1109/ICDM.2017.79},
      • url = {https://www.merl.com/publications/TR2017-168}
      • }
  • MERL Contact:
  • Research Area:

    Data Analytics

Since their introduction over a decade ago, time series motifs have become a fundamental tool for time series analytics, finding diverse uses in dozens of domains. In this work we introduce Time Series Chains, which are related to, but distinct from, time series motifs. Informally, time series chains are a temporally ordered set of subsequence patterns, such that each pattern is similar to the pattern that preceded it, but the first and last patterns are arbitrarily dissimilar. In the discrete space, this is similar to extracting the text chain "hit, hot, dot, dog" from a paragraph. The first and last words have nothing in common, yet they are connected by a chain of words with a small mutual difference. Time series chains can capture the evolution of systems, and help predict the future. As such, they potentially have implications for prognostics. In this work, we introduce a robust definition of time series chains, and a scalable algorithm that allows us to discover them in massive datasets.

 

  • Related News & Events

    •  AWARD   Best Student Paper Award at the International Conference on Data Mining
      Date: November 30, 2017
      Awarded to: Yan Zhu, Makoto Imamura, Daniel Nikovski, Eamonn Keogh
      MERL Contact: Daniel Nikovski
      Research Area: Data Analytics
      Brief
      • Yan Zhu, a former MERL intern from the University of California at Riverside has won the Best Student Paper Award at the International Conference on Data Mining in 2017, for her work on time series chains, a novel primitive for time series analysis. The work was done in collaboration with Makoto Imamura, formerly at Information Technology Center/AI Department, and currently a professor at Tokai University in Tokyo, Japan, Daniel Nikovski from MERL, and Yan's advisor, Prof. Eamonn Keogh from UC Riverside, whose lab has had a long and fruitful collaboration with MERL and Mitsubishi Electric.
    •  
  • Related Publication

  •  Zhu, Y., Imamura, M., Nikovski, D.N., Keogh, E., "Introducing Time Series Chains: A New Primitive for Time Series Data Mining", Knowledge and Information Systems, DOI: 10.1007/s10115-018-1224-8, Vol. 60, No. 2, pp. 1135-1161, August 2019.
    BibTeX TR2019-077 PDF
    • @article{Zhu2019aug,
    • author = {Zhu, Yan and Imamura, Makoto and Nikovski, Daniel N. and Keogh, Eamonn},
    • title = {Introducing Time Series Chains: A New Primitive for Time Series Data Mining},
    • journal = {Knowledge and Information Systems},
    • year = 2019,
    • volume = 60,
    • number = 2,
    • pages = {1135--1161},
    • month = aug,
    • doi = {10.1007/s10115-018-1224-8},
    • url = {https://www.merl.com/publications/TR2019-077}
    • }