Comparative evaluation of a variety of clustering methods on real and simulated data indicates that the appropriate method for a given set of data must be determined empirically. Selection of an appropriate method generally requires several preliminary analyses. With larger data sets, preliminary analyses on the whole may not be possible. As an alternative one may adopt an interactive strategy and break a large set into manageable subsets.
Get full access to this article
View all access options for this article.
References
1.
AnderbergM. R.Cluster analysis for applications. New York: Academic Press, 1973.
BlashfieldR. K.AldenderferM. S.The literature on cluster analysis. Multivariate Behavioral Research, 1978, 13, 271–295.
4.
ByrneD. G.A note on cluster analysis and depression. Australian and New Zealand Journal of Psychiatry, 1978, 12, 99–102.
5.
CarpenterW. T.BartkoJ. J.CarpenterC. L.StraussJ. S.Another view of schizophrenia subtypes. Archives of General Psychiatry, 1976, 33, 508516.
6.
CarrollJ. D.ArabieP.Multidimensional scaling. In RosenzweigM. R.PorterL. W. (Eds.), Annual Review of Psychology. Palo Alto: Annual Review, 1980, in press.
7.
CormackR. M.A review of classification. Journal of the Royal Statistical Society. (Series A), 1971, 134, 321–367.
8.
DixonW. J. (Ed.) BMDP biomedical computer programs. Berkeley: Univer. of California Press, 1975.
9.
DudaR. O.HartP. E.Pattern classification and scene analysis. New York: Wiley, 1973.
HartiganJ.Clustering algorithms. New York: Wiley, 1975.
12.
JarvisR. A.PatrickE. A.Clustering using a similarity measure based on shared nearest neighbours. IEEE Transactions on Computers, 1973, C-22, 1025–1034.
13.
JöreskogK. G.Structural analysis of covariance and correlation matrices. Psychometrika, 1978, 43, 443–477.
14.
KruskalJ. B.The relationship between multidimensional scaling and clustering. In Van RyzinJ. (Ed.), Classification and clustering. New York: Academic Press, 1977. Pp. 17–44.
15.
LevinsonS. E.RabinerL. R.RosenbergA. E.WilponJ. G.Interactive clustering techniques for selecting speaker-independent reference templates for isolated word recognition. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1979, ASSP–27, 134–141.
16.
MilliganG. W.An examination of the effects of six types of error perturbation on fifteen clustering algorithms. Working paper WPS 78–99. College of Administrative Science, Ohio State Univer., Columbus, Ohio, 1978.
17.
PatrickE. A.Fundamentals of pattern recognition. Englewood Cliffs, N. J.: Prentice-Hall, 1972.
18.
SammonJ. W.A nonlinear mapping for data structure analysis. IEEE Transactions on Computers, 1969. C-18, 401–409.
19.
SonquistJ. A.MorganJ. N.The detection of interaction effects. Ann Arbor: Survey Research Center, Institute for Social Research, Univer. of Michigan, 1964.
20.
StraussJ. S.BartkoJ. J.CarpenterW. T.The use of clustering techniques for the classification of psychiatric patients. British Journal of Psychiatry, 1973, 122, 531–540.
