- 478 Downloads
In cluster analysis the multivariate observations are to be divided into g groups. The membership of the groups is not known, nor is the number of groups. The situation is seemingly different from that of discriminant analysis considered in Chapter 6 where both the number of groups and group membership are known. However, there is much in common between our procedure for clustering and the methods we used in the earlier chapters.
KeywordsMahalanobis Distance Subset Size Standardize Distance Forward Search Disperse Group
Unable to display preview. Download preview PDF.