Addressing multicollinearity in semiconductor manufacturing
- 14 January 2011
- journal article
- case study
- Published by Wiley in Quality and Reliability Engineering International
- Vol. 27 (6) , 843-854
Abstract
When building prediction models in the semiconductor environment, many variables, such as input/output variables, have causal relationships which may lead to multicollinearity. There are several approaches to address multicollinearity: variable elimination, orthogonal transformation, and adoption of biased estimates. This paper reviews these methods with respect to an application that has a structure more complex than simple pairwise correlations. We also present two algorithmic variable elimination approaches and compare their performance with that of the existing principal component regression and ridge regression approaches in terms of residual mean square and R2. Copyright © 2011 John Wiley & Sons, Ltd.Keywords
This publication has 22 references indexed in Scilit:
- A Caution Regarding Rules of Thumb for Variance Inflation FactorsQuality & Quantity, 2007
- Regularization and Variable Selection Via the Elastic NetJournal of the Royal Statistical Society Series B: Statistical Methodology, 2005
- Variable Selection in Regression via Repeated Data SplittingJournal of Computational and Graphical Statistics, 1997
- Regression Analysis and Problems of MulticollinearityCommunications in Statistics - Simulation and Computation, 1975
- Latent Root Regression AnalysisTechnometrics, 1974
- Algorithm AS 79: Gram-Schmidt RegressionJournal of the Royal Statistical Society Series C: Applied Statistics, 1974
- Generalized Inverses, Ridge Regression, Biased Linear Estimation, and Nonlinear EstimationTechnometrics, 1970
- Ridge Regression: Biased Estimation for Nonorthogonal ProblemsTechnometrics, 1970
- Ridge Regression: Biased Estimation for Nonorthogonal ProblemsTechnometrics, 1970
- Principal Components Regression in Exploratory Statistical ResearchJournal of the American Statistical Association, 1965