COMPSTAT 2016: Start Registration
View Submission - COMPSTAT
A0426
Title: A new EWMA chart for monitoring the covariance matrix of a multivariate process based on dissimilarity index Authors:  Longcheen Huwang - National Tsing Hua University (Taiwan) [presenting]
Abstract: An EWMA chart is proposed for monitoring the covariance matrix of a multivariate process based on the dissimilarity index of two matrices. The conventional charts for monitoring the covariance matrix of a multivariate process are either based on comparing the sum or the product or both of the eigenvalues of the estimated covariance matrix with those of the in-control covariance matrix. In contrast, the proposed new chart essentially monitor the covariance matrix by comparing the individual eigenvalues of the estimated covariance matrix with those of the in-control counterpart. We compare the performance of the proposed chart with that of the best existing chart in the multivariate normal process. Simulation results show that the proposed EMMA chart outperforms the best existing multivariate EWMA chart for monitoring the covariance matrix. Further, to guarantee that the actual in-control average run length of the proposed chart is not less than the nominal one with a certain probability, we use a bootstrap re-sampling method to adjust the control limit of the proposed chart. Finally, we use an example to demonstrate the applicability of the proposed chart.