Partitional clustering algorithms for highly similar and sparseness Y-Short Tandem Repeat Data / Ali Seman

Clustering is an overlapping method found in many areas such as data mining, machine learning, pattern recognition, bioinformatics and information retrieval. The goal of clustering is to group any similar objects into a cluster, while the other objects that are not similar in the different clusters. Meanwhile, Y-Short Tandem Repeats (Y-STR) is the tandem repeats on Y-Chromosome. The Y-STR data is now being utilized for distinguishing lineages and their relationships applied in many applications such as genetic genealogy, forensic genetic and anthropological genetic applications. This research tends to partition the Y-STR data into groups of similar genetic distances. The genetic distance is measured by comparing the allele values and their modal haplotypes. Nevertheless, the distances among the Y-STR data are typically found similar or very similar to each other. They are characterized by the higher degree of similarity of objects in intra-classes and also inter-classes. In some cases, they are quite distant and sparseness…