A Neura l Network Approach to Synthetic Control Chart for the Process Mean
Loading...
Date
2010-09-30
Authors
Michael, Khoo Boon Chong
Journal Title
Journal ISSN
Volume Title
Publisher
Universiti Sains Malaysia
Abstract
(An abstract of between 100 and 200 words must be prepared in Bahasa Malaysia and in English.
This abstract will be included in the Annual Report of the Research and Innovation Section at a later date
as a means of presenting the project findings of the researcher/s to the University and the community at
large)
In this project, a multivariate synthetic control chart for monitoring the process mean vector of skewed
populations using weighted standard deviations has been proposed. The proposed chart incorporates the
weighted standard deviation (WSD) method of Chang and Bai (2004) into the standard multivariate
synthetic chart of Ghute and Shirke (2008). The standard multivariate synthetic chart consists of the
Hotelling's f2 chart and the conforming run length (CRL) chart. The results show that the proposed chart
performs better than the existing multivariate charts for skewed populations, in terms of false alarm rates
and moderate and large mean shift detection rates based on the various degrees of skewnesses. This
project also suggest a synthetic double sampling (DS) chart for the mean. The synthetic DS chart
comprises a DS sub-chart and a CRL sub-chart. The DS chaIt was proposed by Daudin (1992) while the
CRL chaIt was suggested by Bourke (1991). The synthetic DS chart outperforms both of its standard
counterparts, namely the synthetic chart of Wu and Spedding (2000) and the DS chart of Daudin (1992).
In addition, a comparison between statistical charts and neural network based charts is made, where the
results indicate that neural network based charts perform better.