Enhancing hierarchical clustering with local search

2007

Data Clustering is defined as grouping together objects which share similar properties. These properties can be anything as long as it is possible to measure and compare them. Clustering can be an important tool in many different settings varying from medical use to data mining. In this work we distinguish between two different types of clustering. The simplest one, called partitional clustering, tries to create one solution by comparing objects and partitioning them into non-overlapping groups. The second one, called hierarchical clustering, is a bit more complex and will try to generate a multi-layered solution instead. In this multi-layered structure, groups actually consist of more than…