ABSTRACT
6.1 Introduction
The clustering problem has broad appeal and usefulness as one of the steps in exploratory data analysis [238]. It is an important task in several data mining applications including document retrieval, image/spatial data segmentation, and market analysis [238]. Data mining applications place the following two primary requirements on clustering algorithms: scalability to large data sets (or, the issue of computation time) [29] and non-presumption of any canonical data properties like convexity.
