Journal of the Korean Data and Information Science Society

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

Numerical Investigations in Choosing the Number of Principal Components in Principal Component Regression - CASE II

  • Published : 1999.04.30
Download PDF
⟨ Previous Next ⟩

Abstract

We propose a cross-validatory method for the choice of the number of principal components in principal component regression based on the magnitudes of correlations with y. There are two different manners in choosing principal components, one is the order of eigenvalues(Shin and Moon, 1997) and the other is that of correlations with y. We apply our method to various data sets and compare results of those two methods.

Keywords

References

  1. Technometrics v.13 Mean square error of predictings as a criterion for selecting variables Allen, D.M.
  2. Principal Component Analysis Jolliffe, I.T.
  3. Journal of American Statistical Association v.62 An appraisal of least squares programs of the electionic computer from the point of view of the user Longley, J.W.
  4. Journal of American Statistical Association v.60 Principal components regression in exploratory statistical research Massy, W.F.
  5. Journal of Statistical Theory & Methods v.8 no.2 Numerical investigations in choosing the number of principal components in principal component regression-Case Ⅰ Shin, J.K.;Moon, S.H.
  6. Journal of the uapanese Society of Computational Statistics v.9 Cross-validatory choice for the number of principal components in principal component regression Shin, J.K.;Tanaka, Y.
  7. Bulletin of the Biometric Society of Japan v.10 Sensitvity analysis in principal component regression Shin, J.K.;Tarumi, T.;Tanaka, Y.
  8. Communication in Statiststics : Theory and Methods v.17 Sensitivity analysis in principal component analysis : Influence on the subspace spanned by principal components Tanaka, Y.
  9. Communication in Statistics : Theory and Methods v.18 Influence functions related to eigenvalue problems Which appear in multivariate methods Tanaka, Y.
  10. Correctopms, A v.18 Sensitivity analysis in principal component analysis : Influence on the subspace spanned by principal components Tanaka, Y.