The modern world has witnessed a surge in technological advancements that span various industries. In some sectors, such as search engines, bioinformatics, and pattern recognition, software applications typically deal with having to interpret shear amounts of data in an attempt to discover patterns that may provide great value for business analysis, development, and planning. This emphasized the importance of fields of study such as clustering, a descendant discipline of data mining, which gained momentum in recent decades. Clustering addresses this very problem of analyzing large datasets and attempting to unravel data distributions and patterns by means of a mostly unsupervised data classification [9]. Example clustering applications include multimedia analysis and retrieval [10], pattern recognition [15], and bioinformatics [5].

This chapter starts by providing an overview of existing clustering approaches. Then, it defines key concepts that are utilized by the PYRAMID algorithm. It also presents the experiments that were conducted in Tout et al. [23] as well as other experiments using various datasets that were employed in Sheikholeslami et al. [21] featuring different challenges. Finally, it explores the independence of PYRAMID on user-supplied parameters and outlines future research directions.