Advertisement

Journal of Geodesy

, Volume 83, Issue 9, pp 877–891 | Cite as

Singular value decomposition and cluster analysis as regression diagnostics tools for geodetic applications

  • Markus VennebuschEmail author
  • Axel Nothnagel
  • Hansjörg Kutterer
Original Article

Abstract

It is well known that high-leverage observations significantly affect the estimation of parameters. In geodetic literature, mainly redundancy numbers are used for the detection of single high-leverage observations or of single redundant observations. In this paper a further objective method for the detection of groups of important and less important (and thus redundant) observations is developed. In addition, the parameters which are predominantly affected by these groups of observations are identified. This method thus complements other diagnostics tools, such as, e.g., multiple row diagnostics methods as described in statistical literature (see, e.g., Belsley et al. in Regression diagnostics: identifying influential data and sources of collinearity. Wiley, New York, 1980). The method proposed in this paper is based on geometric aspects of adjustment theory and uses the singular value decomposition of the design matrix of an adjustment problem together with cluster analysis methods for regression diagnostics. It can be applied to any geodetic adjustment problem and can be used for the detection of (groups of) observations that significantly affect the estimated parameters or that are of negligible impact. One of the advantages of the proposed method is the improvement of the reliability of observation plans and thus the reduction of the impact of individual observations (and outliers) on the estimated parameters. This is of particular importance for the very long baseline interferometry technique which serves as an application example of the regression diagnostics tool.

Keywords

Geometry of least-squares adjustment Singular value decomposition Cluster analysis Regression diagnostics Influential data Geodetic VLBI 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Aster R, Borchers B, Thurber C (2005) Parameter estimation and inverse problems. Elsevier, San DiegoGoogle Scholar
  2. Belsley D, Kuh E, Welsch R (1980) Regression diagnostics: identifying influential data and sources of collinearity. Wiley, New YorkGoogle Scholar
  3. Brouwer F (1985) On the principles, assumptions and methods of geodetic very long baseline interferometry, vol 7, Number 4. Netherlands Geodetic Commission, DelftGoogle Scholar
  4. Chatterjee S, Hadi AS (1988) Sensitivity analysis in linear regression. Wiley, New YorkCrossRefGoogle Scholar
  5. Chuvieco E, Congalton RG (1988) Using cluster analysis to improve the selection of training statistics in classifying remotely sensed data. Photogramm Eng Remote Sens vol. 54: 1275–1281Google Scholar
  6. Cook R, Weisberg S (1982) Residuals and influence in regression. Chapman and Hall, New YorkGoogle Scholar
  7. Dermanis A, Rummel R (2000) Data analysis methods in geodesy. In: Dermanis A, Grün A, Sansò F (eds) Geomatic methods for the analysis of data in the Earth sciences. Lecture notes in Earth sciences, vol 95. Springer, Berlin, pp 17–92Google Scholar
  8. Duda R, Hart P, Stork D (2000) Pattern classification. Wiley, New YorkGoogle Scholar
  9. Eeg J (1986) On the adjustment of observations in the presense of blunders. Geodaetisk Institut, Technical Report No. 1, KobenhavnGoogle Scholar
  10. Förstner W (1987) Reliability analysis of parameter estimation in linear models with applications to mensuration problems in computer vision. Comput Vis Graph Image Process 40: 273–310CrossRefGoogle Scholar
  11. Golub G, Kahan W (1965) Calculating the singular values and pseudo-inverse of a matrix. SIAM J Numer Anal 2: 205–224Google Scholar
  12. Grafarend E, Schaffrin B (1993) Ausgleichsrechnung in linearen Modellen. Bibl. Inst. 1993, MannheimGoogle Scholar
  13. Gray J, Ling R (1984) K-clustering as a detection tool for influential subsets in regression. Technometrics 26(4): 305–318CrossRefGoogle Scholar
  14. Hoaglin D, Welsch R (1978) The hat matrix in regression and ANOVA. Am Stat 32(1): 17–22CrossRefGoogle Scholar
  15. Hofmann AD, Maas HG, Streilein A (2003) Derivation of roof types by cluster analysis in parameter spaces or airborne laserscanner point clouds. ISPRS Com.III/3 Workshop ‘3D reconstruction from airborne laserscanner and InSar data’, Dresden, Germany, 8–10 October 2003Google Scholar
  16. Kalman D (1996) A singularly valuable decomposition: the SVD of a matrix. College Math J 27(1): 2–23CrossRefGoogle Scholar
  17. Koch K (1999) Parameter estimation and hypothesis testing in linear models. Springer, BerlinGoogle Scholar
  18. Kovalerchuk B, Schwing J (2004) Visual and spatial analysis: advances in data mining, reasoning, and problem solving. Springer, BerlinGoogle Scholar
  19. Lay D (2003) Linear algebra and its applications. Addison-Wesley, New YorkGoogle Scholar
  20. Leick A (1990) GPS satellite surveying. Wiley, New YorkGoogle Scholar
  21. Lundqvist G (1984) Radio interferometry as a probe of tectonic plate motion. Ph.D. thesis, Chalmers University of Technology, GöteborgGoogle Scholar
  22. Ma C (1978) Very long baseline interferometry applied to polar motion, relativity and geodesy. NASA Technical Memorandum 79582, University of MarylandGoogle Scholar
  23. Menke W (1984) Geophysical data analysis: discrete inverse theory. Academic Press, OrlandoGoogle Scholar
  24. Meyer C (2000) Matrix analysis and applied linear algebra. Society for Industrial and Applied Mathematics (SIAM), PhiladelphiaGoogle Scholar
  25. Nothnagel A (1991) Radiointerferometrische Beobachtungen zur Bestimmung der Polbewegung unter Benutzung langer Nord-Süd-Basislinien, DGK Reihe C, Heft 368, German Geodetic Commission, MunichGoogle Scholar
  26. Parker R (1994) Geophysical inverse theory. Princeton University Press, PrincetonGoogle Scholar
  27. Press W, Plannery B, Teukolsky S, Vetterling W (1988) Numerical recipes. Cambridge University Press, CambridgeGoogle Scholar
  28. Romesburg H (2004) Cluster analysis for researchers. Lulu Press, North Carolina. http://www.lulu.com. Accessed June 2007
  29. Scales J, Smith M, Treitel S (2001) Introductory geophysical inverse theory. Samizdat Press, Golden. ftp://samizdat.mines.edu. Accessed June 2007
  30. Schaffrin B (1997) Reliability measures for correlated observations. J Surv Eng. August:126–137Google Scholar
  31. Schaffrin B (1985) Network design. In: Grafarend E, Sansò F(eds) Optimization and design of geodetic networks. Springer, Berlin, pp 548–597Google Scholar
  32. Schuh, H. et al. (2006) IVS-WG3 report on data analysis, IVS Memorandum 2006-006v01. ftp://ivscc.gsfc.nasa.gov/pub/memos/ivs-2006-006v01.pdf. Accessed June 2007
  33. Shapiro II, Robertson DS, Knight CA, Counselman CC, Rogers AEE, Hinteregger HF, Lippincott S, Whitney AR, Clark TA, Niell AE, Spitzmesser DJ (1974) Transcontinental baseline and the rotation of the earth measured by radio interferometry. Science 186: 920–922CrossRefGoogle Scholar
  34. Sovers O, Fanselow J, Jacobs C (1998) Astrometry and geodesy with radio interferometry: experiments, models, results. Rev Mod Phys 70(4): 1393–1454CrossRefGoogle Scholar
  35. Snieder R, Trampert J (2000) Linear and nonlinear inverse problems. In: Dermanis A, Grün A, Sansò F (eds) Geomatic methods for the analysis of data in the earth sciences. Springer, Berlin, pp 93–120Google Scholar
  36. Stewart G (1993) On the early history of the singular value decomposition. SIAM Rev 35(4):551–566 (ISSN 0036-1445)Google Scholar
  37. Strang G (2003) Lineare algebra. Springer, BerlinGoogle Scholar
  38. Strang G, Borre K (1997) Linear algebra, geodesy and GPS. Wellesley-Cambridge Press, WellesleyGoogle Scholar
  39. Teunissen P (2003) Adjustment theory. An Introduction, Vereniging voor Studie- en Studentenbelangen te Delft (VSSD), DelftGoogle Scholar
  40. Toutenburg H (2003) Lineare modelle. Physica-Verlag, HeidelbergGoogle Scholar
  41. Walter H (1973) Astrometrical applications of long baseline interferometry. Bulletin de la Groupe des Récherches de Géodésie Spatiale (GRGS) No. 10, ToulouseGoogle Scholar
  42. Wang F (2006) Quantitative methods and applications in GIS. CRC Press, Boca RatonGoogle Scholar
  43. Welsch W (1975) Über eine allgemeine reduzierende Gewichtsmatrix zur Elimination von Orientierungsunbekannten. Zeitschrift für Vermessungswesen 100(2): 83–86Google Scholar

Copyright information

© Springer-Verlag 2009

Authors and Affiliations

  • Markus Vennebusch
    • 1
    • 2
    Email author
  • Axel Nothnagel
    • 1
  • Hansjörg Kutterer
    • 3
  1. 1.Institute of Geodesy and Geoinformation of the University of BonnBonnGermany
  2. 2.Institut für Erdmessung, Leibniz University HannoverHannoverGermany
  3. 3.Geodetic Institute, Leibniz University HannoverHannoverGermany

Personalised recommendations