Clustering Data Without Distance Functions.
G. D. S. Ramkumar, Arun N. Swami · 1998
Data mining is being applied with profit in many applications. Clustering or segmentation of data is an important data mining application. One of the problems with traditional clustering methods is that they require the analyst to define distance functions that are not always available. In this paper, we describe a new method for clustering without distance functions. 1 Introduction Mining for information from databases has several important applications [11]. Three of the most common methods to mine data are association rules [1, 2], classification [7], and clustering [8, 3]. Association rules derive patterns from grouped data attributes that co-occur with high frequency. Classification methods produce hierarchical decision models for input data that is sub-divided into classes. Finally, clustering methods group together the records of a data set into disjoint sets that are similar in some respect. Clustering also attempts to place dissimilar records in different partitions. Clusteri...