Journal of applied mathematics & informatics

  • 제4권2호
  • /
  • Pages.501-512
  • /
  • 1997
  • /
  • 2734-1194(pISSN)
  • /
  • 2234-8417(eISSN)
한국전산응용수학회 (The Korean Society for Computational and Applied Mathematics)
권호 표지

DETECTION OF OUTLIERS IN WEIGHTED LEAST SQUARES REGRESSION

  • Shon, Bang-Yong (Department of Computer Engineering Daejin University) ;
  • Kim, Guk-Boh (Department of Computer Engineering Daejin University)
  • 발행 : 1997.06.01

⟨ 이전 논문 다음 논문 ⟩

초록

In multiple linear regression model we have presupposed assumptions (independence normality variance homogeneity and so on) on error term. When case weights are given because of variance heterogeneity we can estimate efficiently regression parameter using weighted least squares estimator. Unfortunately this estimator is sen-sitive to outliers like ordinary least squares estimator. Thus in this paper we proposed some statistics for detection of outliers in weighted least squares regression.

키워드

참고문헌

  1. Journal of Royal Statistical Society v.B44 Regression diagnostics, transformations and constructed variables Atkinson, A.C.
  2. Statistical Science v.1 Influential observations, high leverage points, and outlier in linear regression Chatterjee, S.;Hadi, A.S.
  3. Residuals and Influence in Regression Cook, R.D.;Weisberg, S.
  4. Journal of American Statistical Association v.68 The joint distribution of the studentized least squares residulas from a general linear regression Ellenberg, J.H.
  5. American Statistician v.32 The hat matrix in regression and ANOVA Hoaglin, D.C.;Welsch, R.
  6. Robust regression in Exploring Data Tables, Tredns and Shapes Li, G.;D.C. Hoaglin(ed.);F. Mosteller(ed.);J.W. Turkey(ed.)
  7. Journal of American Statistical Association v.85 Unmasking multivariate outliers and leverage points Ruosseeuw, P.J.;Zomerenm, B.C.
  8. Doctoral Thesis Weighted Least Squares Regression Diagnostics and its Applicatin to Robust Regression Sohn, B.Y.
  9. Sanhky a A v.23 Testing for a single outlier in a regression model Srikantan, K.S.