Methods and Algorithms for Robust Filtering
We discuss filtering procedures for robust extraction of a signal from noisy time series. Moving averages and running medians are standard methods for this, but they have shortcomings when large spikes (outliers) respectively trends occur. Modified trimmed means and linear median hybrid filters combine advantages of both approaches, but they do not completely overcome the difficulties. Improvements can be achieved by using robust regression methods, which work even in real time because of increased computational power and faster algorithms. Extending recent work we present filters for robust online signal extraction and discuss their merits for preserving trends, abrupt shifts and extremes and for the removal of spikes.
Key wordsSignal extraction drift edge outlier update algorithm
Unable to display preview. Download preview PDF.
- Cormen T.H., Leiserson C.E., Rivest R.L. (1990). Introduction to algo-rithms. MIT Press, Cambridge, Massachusetts, and McGraw-Hill Book Company, New York.Google Scholar
- Croux C., Rousseeuw P.J. (1992). Time-efficient algorithms for two highly robust estimators of scale. COMPSTAT 1992, Physica-Verlag, Heidelberg, 411–428.Google Scholar
- Fried R. (2004). Robust filtering of time series with trends. Current Advances and Trends in Nonparametric Statistics, special issue of the J. of Nonparametric Statistics, to appear.Google Scholar
- Fried R., Bernholt T., Gather U. (2004a). Repeated median and hybrid filters. Technical Report, SFB 475, University of Dortmund, Germany.Google Scholar
- Fried R., Bernholt T., Gather U. (2004b). Modified repeated median filters. Preprint, Department of Statistics, University of Dortmund, Germany.Google Scholar
- Fried R., Gather U. (2002). Fast and robust filtering of time series with trends. COMPSTAT 2002, Physica-Verlag, Heidelberg, 367–372.Google Scholar
- Tukey J.W. (1977). Exploratory data analysis. Addison-Wesley, Reading, Mass, (preliminary ed. 1971).Google Scholar