Introduction to Pattern Recognition: A Matlab Approach

7.7 Hierarchical Clustering Algorithms 203

To run the generalized agglomerative scheme (GAS), type

[bel, thres] = agglom(prox_mat,code)

where

prox_mat is the N×N dissimilarity matrix for the N vectors of the data set X,

( prox_mat(i, j) is the distance between vectors x

i

and x

j

),

code is an integer indicating the specific clustering algorithm to be used (1 stands for single link;

2 stands for complete link),

bel is an N ×N matrix whose (i, j) element contains the cluster label for the jth vector in the ith

clustering. (The first row of bel corresponds to the N -cluster clustering; the second row, to the

(N −1)-cluster clustering; and the Nth row, to th