Journal of Geodesy

, Volume 81, Issue 2, pp 137–148 | Cite as

Effect of heteroscedasticity and heterogeneousness on outlier detection for geodetic networks

Original Article

Abstract

The robustness of an outlier detection method strongly depends on the weights of observations, i.e., the type of the stochastic model applied (homoscedasticity, heteroscedasticity and heterogeneousness). In this paper, we have investigated how the reliability of the robust methods and tests for outliers changes depending on the weights of the observations in geodetic networks. Furthermore, the contribution of the directions and distances to horizontal control network with regard to reliability are investigated separately. The concept of a breakdown point is used as a global measure of robustness against outliers. The mean success rate (MSR) is found to be a practical tool for confirming the breakdown point. Many different “good” data samples are generated for each network and then deliberately contaminated using a Monte-Carlo simulation. Six robust methods and Baarda’s test are applied to the corrupted samples and the degree of corruption is varied. The performance of each method is measured using both local and global MSRs. Our research shows: (1) The MSRs of Baarda’s test change depending on the strength of the heteroscedasticity, but do not change for trilateration and leveling networks, (2) the global MSRs of robust methods do not differ considerably from the local ones

Keywords

Robust estimation Reliability Outlier detection Heteroscedasticity heterogeneousness Global and local mean success rates (MSRs) 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Awange JL, Aduol FWO (1999). An evalution of some robust estimation techniques in the estimation of geodetic parameters. Surv Rev 35 (273):146–162Google Scholar
  2. Andrews D (1974). A robust method for multiple linear regression. Technometrics 16:523–531CrossRefGoogle Scholar
  3. Baarda W (1968) A testing procedure for use in geodetic networks. Publication on Geodesy. New series 2, no. 5. Netherlands Geodetic Commission, DelftGoogle Scholar
  4. Barrodale I, Roberts FDK (1974). Solution of an over determined system of equations in L 1-norm. Commun ACM 17:319–320CrossRefGoogle Scholar
  5. Benning W (1995). Vergleich dreier Lp-Schaetzer zur Fehlersuche in hybriden Lagenetzen. Zeitschrift fuer Vermessungswesen 12:606–617Google Scholar
  6. Berber M, Hekimoglu S (2001) What is the reliability of robust estimators in networks? In: 1st International Symposium on Robust Statistics and Fuzzy Techniques in Geodesy and GIS March 12–16, ETH, Zurich, pp 61–66Google Scholar
  7. Bickel PJ (1978). Using residuals robustly I: Tests for heteroscedasticity, nonlinearity. Ann Stat 6(2):266–291Google Scholar
  8. Box GEP, Hill WJ (1974). Correcting inhomogeneity of variance with power transformation weighting. Technometrics 16:385–389CrossRefGoogle Scholar
  9. Carroll RJ, Ruppert D (1982). Robust estimation in heteroscedastic linear models. Ann Stat 10(2):429–441Google Scholar
  10. Donoho DL, Huber PJ (1983) The notion of breakdown point. In: Bickel PJ, Doksum KA, Hodges JL Jr (eds) A Festschrift for Erich L. Lehmann Wadsworth, Belmont, pp 157–184Google Scholar
  11. Erenoglu RC, Hekimoglu S (2004) Effects of gross errors on reliability and datum in geodetic networks. In: International Symposium on Modern Technologies, Education and Professional Practice in Geodesy and Related Fields, Sofia, 04–05 November 2004, pp 465–470Google Scholar
  12. Fuchs H (1982). Contribution to the adjustment by minimizing the sum of absolute residuals. Manuscr Geod 7:151–207Google Scholar
  13. Gao Y, Krakinwsky EJ, Czompo J (1992). Robust testing procedure for detection of multiple blunders. J Surv Eng 118(1):11–23Google Scholar
  14. Hampel F, Ronchetti E, Rousseeuw P, Stahel W (1986). Robust statistics: the approach based on influence functions. Wiley, New YorkGoogle Scholar
  15. Harvey BR (1993). Survey network adjustments by the 1 method. Aust J Geod Photogram Surv 59:39–52Google Scholar
  16. Hekimoglu S (1997). The finite sample breakdown points of the conventional iterative outlier detection procedures. J Surv Eng 123(1):15–31CrossRefGoogle Scholar
  17. Hekimoglu S (2005). Do robust methods identify outliers more reliably than conventional tests for outliers? Zeitschrift fuer Vermessungswesen (05/03):174–180Google Scholar
  18. Hekimoglu S, Berber M (2001) Reliability of robust methods in heteroscedastic linear models. IAG 2001, Scientific Assembly, Budapest, 2–7 September 2001, abstract book, p 53Google Scholar
  19. Hekimoglu S, Berber M (2003). Effectiveness of robust methods in heterogeneous linear models. J Geod 76(11–12):706–713CrossRefGoogle Scholar
  20. Hekimoglu S, Koch KR (1999) How can reliability of the robust methods be measured? In: Altan MO, Gründig L (eds) Third Turkish–German joint geodetic days, Istanbul, 1–4 June, 1999, pp 179–196Google Scholar
  21. Hekimoglu S, Koch KR (2000). How can reliability of tests for outliers be measured? Allg Vermessungs-Nachr 107(7):247–254Google Scholar
  22. Hekimoglu S, Sanli U (2003). Changing residuals by reweighting one or more observations. Zeitschrift fuer Vermessungswesen (04/03):271–277Google Scholar
  23. Huber PJ (1964). Robust estimation of a location parameter. Ann Math Stat 35:73–101Google Scholar
  24. Huber PJ (1981). Robust statistics. Wiley, New YorkGoogle Scholar
  25. Kampmann G (1989). Zur Ausgleichung freier Netze mit der L 1-norm-methode. Allg Vermessungs-Nachr 96:110–118Google Scholar
  26. Koch KR (1996). Robuste Parameterschaetzung. Allg Vermessungs-Nachr 103:1–18Google Scholar
  27. Koch KR (1999). Parameter estimation and hypothesis testing in linear models, 2nd edn. Springer, Berlin Heidelberg New YorkGoogle Scholar
  28. Krarup T, Juhl J, Kubik K (1980) Götterdaemmerung over least squares adjustment. In: Procedings 14th congress of int. soc. photogr. HamburgGoogle Scholar
  29. Marchall J (2002). L 1-norm pre-analysis measures for geodetic networks. J Geod 76:345–352CrossRefGoogle Scholar
  30. Rousseeuw PJ, Leroy AM (1987). Robust regression and outlier detection. Wiley, New YorkGoogle Scholar
  31. Pope AJ (1976) The statistics of residuals and the outlier detection of outliers. NOAA Technical Report NOS 65 National Geodetic Survey, RockvilleGoogle Scholar
  32. Wicki F (1999) Robuste Schätzverfahren für die parameterschätzung in geodätischen Netzen. Institut für Geodäsie und Photogrammetrie an der ETH, Zürich, Mitt. Nr. 67Google Scholar
  33. Wilcox RR (1997). Introduction to robust estimation and hypothesis testing. Academic, New YorkGoogle Scholar
  34. Xu PL (1989). On the robust estimation with correlated observations. Bull Geod 63:237–252CrossRefGoogle Scholar
  35. Xu PL (1993). Consequences of constant parameters and confidence intervals of robust estimation. Boll Geod Sc Aff 52:231–249Google Scholar
  36. Xu PL (2005). Sign-constrained robust least squares, subjective breakdown point and the effect of weights of observations on robustness. J Geod 79:146–159CrossRefGoogle Scholar
  37. Yang Y (1994). Robust estimation for dependent observations. Manuscr Geod 19(1):10–17Google Scholar
  38. Yang Y (1999). Robust estimation of geodetic datum transformation. J Geod 73:268–274CrossRefGoogle Scholar
  39. Youcai H (1995). On the design of estimators with high breakdown points for outlier identification in triangulation networks. Bull Geod 69:292–299CrossRefGoogle Scholar

Copyright information

© Springer-Verlag 2006

Authors and Affiliations

  1. 1.Department of Geodesy and Photogrammetry EngineeringYıldız Technical UniversityYıldız, IstanbulTurkey

Personalised recommendations