An R package for Cluster Validation
Clustering is the task of partitioning a set of objects into clusters of similar objects, while at the same time maximizing the difference between objects in different clusters. In order to be successful, clustering needs quantifiable measures, which evalute how good the clusters are.
The two main groups of measures will be implemented - external and internal validation measures. External validation makes use of information not present in the data to assess whether the formed clusters match some external structre (e.g. pre-existing object labels). On the other hand, internal validation measures evaluate the goodness of a clustering structure without respect to any external information.
This package is not under development. It currently has very few external and only one internal cluster validation measures implemented.
The ev.vanDongen() function computes the value of the van Dongen crietrion. More about it can be read here.
The ev.mi() function calculates the mutual information measure.
The ev.vigneron() is similar to ev.mi() but captures better imbalanced clusters. More about it can be read here.
The iv.xb() function is an internal validation functions which calculates the Xie-Beni index. It can be used to validate fuzzy clustering as well. Two variants are supplied (normal and weighted).