TPCC: Proxy Based Trusted Pre Cluster Count in Unlabelled Data Sets
The selection of the number of clusters is an important and challenging issue in cluster analysis. A number of attempts have been made to estimate the number of clusters in a given dataset. Most methods are post clustering measures of cluster validity i.e. they attempt to choose the best partition from a set of alternative partitions. In contrast, tendency assessment attempts to estimate the number of clusters before clustering occurs. In this paper, the authors investigate a new method called Trusted Pre-Cluster Count (TPCC) algorithm for automatically estimating the number of clusters in unlabeled datasets, which is based on an existing algorithm Dark Block Extraction (DBE) of a dataset, using several common image and signal processing techniques.