Combining Graphical Models and PCA for Statistical Process Control
Principal component analysis (PCA) is frequently used for detection of common structures in multivariate data, e.g. in statistical process control. Critical issues are the choice of the number of principal components and their interpretation. These tasks become even more difficult when dynamic PCA (Brillinger, 1981) CitationRef CitationID Omitted tag 1 bachieve is applied to incorporate dependencies within time series data. We use the information obtained from graphical models to improve pattern detection based on PCA.
KeywordsTime series analysis dimension reduction online monitoring pattern detection
Unable to display preview. Download preview PDF.
- Dahlhaus, R. and Eichler, M. (2000). Spectrum. Program is available at http://www.statlab.uni-heidelberg.de/projects/graphical.models/projects/graphical.modelsGoogle Scholar
- Gather, U., Imhoff, M. and Fried, R. (2002). Graphical models for multivariate time series from intensive care monitoring. Statistics in Medicine, to appear.Google Scholar
- Keller, M. (2000). Hauptkomponentenanalyse für intensivmedizinische Zeitreihen (in German). Unpublished Diploma Thesis, Department of Statistics, University of Dortmund, Germany.Google Scholar
- MacGregor, J.F., Jaeckle, C., Kiparissides, C. and Koutoudi, M. (1994). Process Monitoring and Diagnosis by Multiblock PLS Methods. AIChE J., 40 826–838.Google Scholar