Compstat pp 367-372 | Cite as

Fast and Robust Filtering of Time Series with Trends

  • Roland Fried
  • Ursula Gather


Fast and robust methods are needed for denoising time series data measured with high sampling frequencies. In intensive care e.g. physiological variables like the heart rate are observed in short time intervals. Systematic changes have to be detected quickly and distinguished from clinically irrelevant short term fluctuations and artifacts. Median filtering works well if there is no substantial trend in the data but improvements are possible by approximating the data by a local linear trend.


Signal extraction linear regression level shifts outliers 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Davies, P.L., Fried, R. and Gather, U. (2002). Robust signal extraction for on-line monitoring data. Technical Report 1/2002, SFB 475, University of Dortmund, Germany.Google Scholar
  2. Siegel, A.F. (1982). Robust regression using repeated medians. Biometrika,68, 242–244.CrossRefGoogle Scholar
  3. Sposito, V.A. (1990). Some properties of L p-Estimators. In: K.D. Lawrence and J.L. Arthur (eds.). Robust Regression. Analysis and Applications, 23–58. New York: Marcel Dekker.Google Scholar
  4. Stromberg, A.J. (1993). Computing the exact least median of squares estimate and stability diagnostics in multiple linear regression. SIAM J. Sci. Comput., 14, 1289–1299.zbMATHCrossRefGoogle Scholar
  5. Tukey, J.W. (1977). Exploratory Data Analysis. Reading, Mass.: Addison-Wesley.zbMATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2002

Authors and Affiliations

  • Roland Fried
    • 1
  • Ursula Gather
    • 1
  1. 1.Department of StatisticsUniversity of DortmundDortmundGermany

Personalised recommendations