August 2018
Intermediate to advanced
522 pages
12h 45m
English
Another method that is based on the concept of dense and well separated clusters is the Calinski-Harabasz index. To build it, we first need to define the inter-cluster dispersion. If we have k clusters with their relative centroids and the global centroid, the inter-cluster dispersion or Between Cluster Dispersion (BCD) is defined as follows:

In the preceding expression, ni is the number of elements belonging to the cluster, i, μ is the global centroid, μi is the centroid of cluster i, and Tr(•) is the trace of a square matrix. The intra-cluster dispersion or Within Cluster Dispersion (WCD) is defined as follows:
Read now
Unlock full access