SavedAdd your journal

Journals / Harita Dergisi / 2006 / Cilt: 73 - Sayı: 135

Problems arising in the analysis of outliers

Kaba hataların belirlenmesindeki sorunlar

Şerif HEKİMOĞLU
Pages
80–93
DOI
—
🏛️Bibliographic metadata and full text access for this article are provided via DergiPark.
DergiPark Record ↗
✨

AI Summary & Key Takeaways

AI Insight
⚡ TL;DR(Too Long; Didn't Read)

According to the results of simulation, the mean success rates of robust methods are greater than the ones of the conventional tests for outliers when considering all the cases which problems occur such as leverage point, masking effect and swamping effect etc.

📌 3 Key Takeaways

🎯
Objective & Hypothesis:In order to identify outliers, there are two approaches: the conventional tests for outliers and robust methods.
📊
Key Findings & Data:In order to compare the success of these approaches to identify outliers with each other, mean success rate is used.
💡
Conclusion & Impact:According to the results of simulation, the mean success rates of robust methods are greater than the ones of the conventional tests for outliers when considering all the cases which problems occur such as leverage point, masking effect and swamping effect etc.
🔬Design:Research Article
📍Scope:Comprehensive Sample
⚙️Variables:Domain Parameters

Abstract

In order to identify outliers, there are two approaches: the conventional tests for outliers and robust methods. In this study, it is given information about that the problems arise in the analysis of outliers generally, and how these approaches give solutions to these questions. In order to compare the success of these approaches to identify outliers with each other, mean success rate is used. According to the results of simulation, the mean success rates of robust methods are greater than the ones of the conventional tests for outliers when considering all the cases which problems occur such as leverage point, masking effect and swamping effect etc. Therefore, the robust methods should be used. However, when the observations don't include any outlier, robust methods generate outlier. So, a good observation may appear as a bad one. This is a risk that we must take into the consideration.

Özet

Kaba hataları belirlemek için iki temel yaklaşım bulunmaktadır: Klasik uyuşumsuz ölçü testleri ve robust yöntemler. Bu yazıda kaba hata belirlemede ortaya çıkan genel sorunlar üzerinde bilgi verilmekte ve sorunlar karşısında bu iki yaklaşımın çözümleri tartışılmaktadır. Gerek robust yöntemlerin ve gerekse uyuşumsuz ölçü testlerinin sonuçlarının güvenirliği ortalama başarı oranı ile ölçülmektedir. Yapılan simülasyon çalışmalarına göre, robust yöntemlerin, genel olarak tüm sorunlu durumlar (kaldıraç noktası, gizleme ve batma etkisi, kritik kaba hatalar) gözönüne alındığında, uyuşumsuz ölçü testlerine göre daha başarılı olduğu görülmüştür. Dolayısıyla kaba hatalı ölçüleri belirlemek için robust yöntemler kullanılmalıdır. Buna karşılık, eğer örnek kümede hiç uyuşumsuz ölçü yoksa, robust yöntemler kaba hatalı ölçü üretir, yani iyi bir ölçü kaba hatalıymış gibi görünür. Bu riski gözardı etmemek gerekir.

🔖

Cite This Article

HEKİMOĞLU, Ş. (2006). Problems arising in the analysis of outliers. Harita Dergisi, Cilt: 73 - Sayı: 135, 80–93. https://ajindex.com/articles/201429

References (23)

  1. Aksoy, A. Uyuşumsuz Ölçüler Testi. Harita Dergisi, 93, 1984.
  2. Ayan, T. Uyuşumsuz Ölçüler Testi. Harita ve Kadastro Mühendisliği, Sayı 72, 1992.
  3. Ayhan, E. Aksoy, Z.N. Robust Kestirim ve Kaba Hatalı Ölçülerin Belirlenmesi. Harita Dergisi, Sayı 106, 1991.
  4. Baarda, W. A Testing Procedure for Use in Geodetic Networks. Netherlands Geodetic Com., New Series, Delft, Netherlands, 2(5), 1968
  5. Barrodale, I., Roberts, F.D.K Solution of an Over Determined System of Equations in L1- norm. Comm. ACM, vol. 17: 319-320, 1974.
  6. Chatterjee, S., Hadi, AS. Sensitivity Analysis in Linear Regression. John Wiley&Sons, Inc., New York, 1988.
  7. Donoho, D.L., Huber, P.J The Notion of Breakdown Point. In a Festschrift for Erich Lehmann, P.J.Bickel, K.A.Doksum, and J.L.Hodges, Jr. (Eds), Wadsworth, Belmont, Cal.:157- 184, 1983.
  8. Hadi, A.S., Siminoff, J.S. Procedures for the Identification of Multiple Outliers in Linear Models. J. Am. Stat. Assoc.88: 1264-1272, 1993.
  9. Hampel, F., Ronchetti, E., Rousseeuw, P., Stahel, W. Robust Statistics: The approach based on influence functions. John Wiley and Sons, New York, N.Y., 1986.
  10. He, X., Simpson, D.G., Portnoy, S.L Breakdown Robustness of Tests. J. Am. Statistical. Assn., 85: 446-452, 1990.
  11. Hekimoğlu, Ş. The Finite Sample Breakdown Points of the Conventional İterative Outlier Detection Procedures. J. Surv. Eng., ASCE, 123(1), 15-31, 1997.
  12. Hekimoğlu, Ş. Equiredundancy Design and its Application to MEstimation. J.Surv. Eng., ASCE, 124(3):103-124, 1998.
  13. Hekimoğlu, Ş., Koch, K.R. How Can Reliability of the Robust Methods Be Measured? Third Turkish-German Joint Geodetic Days. 1-4 June 1999,
  14. Istanbul, Vol.1, 179-196, 1999
  15. Hekimoğlu, Ş., Koch, K.R. How Can Reliability of Tests for Outliers Be Measured ? AVN, 107(7): 247-254, 2000.
  16. Hekimoğlu, Ş. Do Robust Methods İdentify Outliers More Reliably Than Conventional Tests For Outliers? Zfv, 130, 3, 174- 180, 2005.
  17. Hekimoğlu, Ş. Increasing Reliability of Data Snooping when Outliers are Small. AVN, 112(1): 7-12, 2005
  18. Hekimoğlu, Ş. Increasing Reliability of the Tests for Outliers Whose Magnitude is Small. Survey Review, 38(298): 274-285, 2005
  19. Huber, P.J. Robust Estimation of a Location Parameter. Annals of Mathematical Statistics, 35: 73-101, 1964.
  20. Huber, P.J. Robust Statistics. John Wiley and Sons, Inc., New York, N.Y., 1981.
  21. Kampmann, G. Robuste Deformationsanalyse mittels balancierter Ausgleichung. AVN, 101: 8-17, 1994.
  22. Koch, K.R. Robuste Parameterschaetzung. AVN, 103:1-18, 1996.
  23. Koch, K.R. Parameter Estimation a