Fabric handle evaluation can be reduced to a typical clustering problem, and fuzzy cluster analysis is applied to a practical example in this paper. For comparison, the same example is also tackled using hierarchical cluster methods of multivariate analysis. Both kinds of results have been found to be consistent. To determine the optimum group number g of clustering, Marriot's g2 |W| criterion is suggested and the result is demonstrated as satisfactory; consequently it will be possible to use discriminant analysis to evaluate fabric handle.
Get full access to this article
View all access options for this article.
References
1.
DuranB. S.OdellP. L., “Cluster Analysis - A Survey,”Springer-Verlag, Berlin, 1974.
2.
EverittB. S., Unresolved Problems in Cluster Analysis, Biometrics35, 169 (1979).
3.
FuK. S.WhinstonA. B., “Pattern Recognition Theory and Application,”Sijthoff and Noordhoff International, The Netherlands, 1977, p. 453.
4.
MarriotF. H. C., Practical Problems in a Method of Cluster Analysis, Biometrics27, 501 (1971).
5.
PanN.YenK. C., A New Measurement Proposal for Fabric Handle Evaluation, J. China Textile Univ. Eng. Ed., no. 2, 1 (1986).
6.
WishartW., An Algorithm for Hierarchical Classification, Biometrics25, 165 (1969).
7.
ZadehL. A., Fuzzy Sets, Inf. Control8, 338 (1965).
8.
ZhangY. Z.FangK. T., “Introduction to Multivariate Statistical Analysis,”Science Press, Beijing, 1982, pp. 393, 401.
