A new approach to hierarchical clustering for the analysis of genomic data

Clustering algorithms in biomedical disciplines are usually selected between two main families, k-means and agglomerative hierarchical clustering. These methods are well studied and well established. However, both categories have some drawbacks related to data dimensionality (for partitional algorithms) and to the bottom-up structure (for hierarchical algorithms). To overcome these limitations, we present a hierarchical clustering algorithm based on a completely different principle, which is the analysis of shared farthest neighbors. The principle of operation and the rationale are illustrated, and experimental results on different data sets are presented.