M. Soskic et al., LINK BETWEEN ORTHOGONAL AND STANDARD MULTIPLE LINEAR-REGRESSION MODELS, Journal of chemical information and computer sciences, 36(4), 1996, pp. 829-832
Citations number
16
Categorie Soggetti
Information Science & Library Science","Computer Application, Chemistry & Engineering","Computer Science Interdisciplinary Applications",Chemistry,"Computer Science Information Systems
Several topics in connection with a recently proposed method for the o
rthogonalization of predictor variables (dominant component analysis)
are considered. Applying the sequential regression procedure, it is sh
own that dominant component analysis and the standard multiple linear
regression method are directly related to each other. In addition, it
is demonstrated that an earlier proposed iterative procedure for the o
rthogonalization of a correlated variable can be efficiently replaced
by one step regression. It is also shown that the coefficient of deter
mination for an orthogonal descriptor coincides with the corresponding
squared semipartial correlation coefficient. Finally, the origin of e
xtra information in an orthogonalized predictor variable is discussed.