Visual Analytics for Large-scale Time Series

Problem

Multivariate time series are common in many sectors and fields such as finance, energy or biology. They are also often long and high-dimensional: As an example, 5 years of intraday stock data for 50 stocks contain 50 dimensions and >2.5M time points. This is too much data to display in its raw form, and often needs to be reduced in some way for visualization. Such approaches could e.g. be clustering of daily patterns, or applying dimension reduction algorithms to time slices.
 

Aim

Provide a systematic overview of how VA applications reduce, if so, and visualize very long or high-dimensional time series.

Other information

References

  • M. Steiger, J. Bernard, S. Mittelstädt, H. Lücke-Tieke, D. Keim, T. May, J. Kohlhammer, “Visual Analysis of Time-Series Similarities for Anomaly Detection in Sensor Networks: Visual Analysis of Time-Series Similarities for Anomaly Detection in Sensor Networks,” Computer Graphics Forum, vol. 33, no. 3, pp. 401–410, Jun. 2014, doi: 10.1111/cgf.12396.
  • D. Jäckle, F. Fischer, T. Schreck, and D. A. Keim, “Temporal MDS Plots for Analysis of Multivariate Data,” IEEE Transactions on Visualization and Computer Graphics, vol. 22, no. 1, pp. 141–150, Jan. 2016, doi: 10.1109/TVCG.2015.2467553.
  • M. Ali, M. W. Jones, X. Xie, and M. Williams, “TimeCluster: dimension reduction applied to temporal data for visual analytics,” Vis Comput, vol. 35, no. 6–8, pp. 1013–1026, Jun. 2019, doi: 10.1007/s00371-019-01673-y.

Contact

Gallery

Further information

Topics
High-dimensional time series, large time series, visual analytics, dimension reduction, data aggregation
Area
Visual Analytics (VA)
English
Scope
SE
Status
open