Modified Regression Technique to Reduce Effects of Multicollinearity
Modified Regression Technique to Reduce Effects of Multicollinearity
Author(s): Slav Angelov, Eugenia StoimenovaSubject(s): Social Sciences, Economy, Education, ICT Information and Communications Technologies
Published by: Нов български университет
Keywords: Prediction; Multicollinearity Correction; Multivariate Linear Regression; Orthogonal Regression; Principal Component Regression (PCR); Partial Least Squares Regression (PLSR);
Summary/Abstract: In this paper an algorithm for transforming correlated regressors from OLS model to sets of orthogonal ones is briefly presented. It is demonstrated that the obtained orthogonal components can be used to create models that reduce the prediction error under multicollinearity. The results are compared with the ones from Principal component regression(PCR) and Partial least squares regression(PLSR) which are similar techniques.
Journal: Computer Science and Education in Computer Science
- Issue Year: 12/2016
- Issue No: 1
- Page Range: 353-367
- Page Count: 15
- Language: English
