Journal of the Korean Data and Information Science Society

The Korean Data and Information Science Society (한국데이터정보과학회)
권호 표지

A Method of Choosing a Value of the Bending Constant in Huber's M-Estimation Function

  • Park, Ro-Jin (Department of Information and Statistics, Taejon University)
  • Published : 2000.10.31
Download PDF
⟨ Previous Next ⟩

Abstract

The shape of an M-estimation function is generally determined in the sense of either/both maximizing efficiency of an M-estimator at the model or/and bounding the influence function of an M-estimator. We propose an empirical method of choosing a value of the bending constant in Huber's ${\psi}-function$, which is the most widely used M-estimation function when estimating the location parameter.

Keywords

References

  1. Technometrics v.19 Detection of outliers in Linear Regression Cook, R. D.
  2. Technometrics v.19 Letter to the editor Cook, R. D.
  3. Communications in Statistics - Theory and Methods v.17 no.3 The Likelihood Displacement: A Unifying Principle for Influence Measures Cook, R. D.;Weisberg, S.
  4. Residuals and Influence in Regression Cook, R. D.;Weisberg, S.
  5. Journal of Physiology v.32 The Action of Optical Isomers Ⅱ, Hyoscines Cushny, A. R.;Peebles, A. R.
  6. Unpublished Ph.D. dissertation Salience of the self and salience of standards: Attempts to match self to standard. Dana, E.
  7. Robust Statistics, the Approach Based on Influence Functions Hampel, F. R.;Ronchetti, E. M.;Rousseeuw, P. J.;Stahel, W. A.
  8. Robust Statistics Huber, P. J.
  9. Robust Estimation and Testing Staudte, R. G.;Sheather, S. J.
  10. Introduction to Robust Estimation and Hypothesis Testing Wilcox, R. R.