Finding Number of Clusters before Finding Clusters

Determining number of clusters present in a data set is an important problem in cluster analysis. Conventional clustering techniques generally assume this parameter to be user supplied. There exist very few techniques that can solve the problem of automatic detection of number of clusters satisfactorily. Some of these techniques rely on user supplied information, while some others use cluster validity indices which are expensive with regard to computation time. A recently developed visual mechanism for determining the clustering tendency (VAT, Visual Assessment of Tendency for clustering) present in a data set has become very popular. We shall show how VAT-based algorithms may be used for automatic determination of number of clusters very efficiently.