Application of Copulas in Multivariate Quality Control Problems

Please enable javascript in your browser.

Bakhshiani, Asghar | 2010

874 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 40600 (01)
University: Sharif University of Technology
Department: Industrial Engineering
Advisor(s): Akhavan Niaki, Taghi
Abstract:
In this research, we consider the application of copulas in multivariate quality control problems. In particular, we consider two specific problems. The first problem concerns the situation where the normality assumption is rejected. In this case, copulas can be used as a flexible tool to define a broad range of multivariate distributions with different dependence structure as well as marginal distributions. The second problem concerns proposing control charts to monitor the dependence structure among quality characteristics. The proposed method not only produces an out-of-control signal when dependence structure among variables deviates from the specified one, but also can be used to interpret the out-of-control signals to specify the variables relationship among them have resulted in the out-of-control state
Keywords:
Multivariate Quality Control ; Control Chart ; Copulas ; Rosenblatt Transformation ; Dependence Structure

No TOC