Addressing multicollinearity in semiconductor manufacturing
Article first published online: 14 JAN 2011
Copyright © 2011 John Wiley & Sons, Ltd.
Quality and Reliability Engineering International
Volume 27, Issue 6, pages 843–854, October 2011
How to Cite
Chang, Y.-C. and Mastrangelo, C. (2011), Addressing multicollinearity in semiconductor manufacturing. Qual. Reliab. Engng. Int., 27: 843–854. doi: 10.1002/qre.1173
- Issue published online: 15 SEP 2011
- Article first published online: 14 JAN 2011
- variable elimination;
- principal components regression;
- variance inflation factor
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.