A LASSO-Based Diagnostic Framework for Multivariate Statistical Process Control
A LASSO-Based Diagnostic Framework for Multivariate Statistical Process Control
- Publication:
- Technometrics
- Date:
- August 2011
- Issue:
- Volume 53 Issue 3
- Pages:
- pp. 297–309
- Author(s):
- Zou, Changliang, Jiang, Wei, Tsung, Fugee
- Organization(s):
- Nankai University, Tianjin, China, Shanghai Jiaotong University, Shanghai, China, Hong Kong University of Science and Technology, Hong Kong, China
Abstract
[This abstract is based on the authors' abstract.]This article proposes a fault isolation procedure combining Bayesian information criterion with the adaptive LASSO variable selection method. The diagnostic result is obtained quickly and with about as much computational effort as a least squares regression, and it requires no additional testing. Several statistical process control examples are presented to demonstrate the effectiveness of the method.