Abstract
This article proposes a methodology that helps to predict the main mean shifts, denoted as principal alarms, in a non-normal multivariate process using the available in-control data. The analysis is based on the transformation of the observed correlated variables into independent factors using independent component analysis. These independent components allow us to simulate shifts preserving the covariance structure. The graphical representations of those simulated shifts are helpful in improving the design and control of the process. Two real manufacturing processes are presented showing the advantage of the proposed methodology.
Acknowledgements
The authors are grateful to Alcalá Indistrial SA for providing the case study, the data, and for permission to use . We are especially grateful to Quality Engineer Jose A. Delgado-Echague for his comments and support. González's research is partly supported by CICYT grant DPI2005-08018. Sánchez's research is partly supported by grant CAM 06/HSE/0174/2004 and CICYT grant SEJ2004-03303. The authors are also grateful to the anonymous referee for insightful comments that significantly improved the manuscript.