The role of conceptual hierarchies in the diagnosis and prevention of diabetes

Clustering is a data mining technique, in which objects of similar characteristics are grouped together to form a cluster. Traditional clustering algorithms use distance metric measures to form clusters out of data which produce unstable results. More over traditional clustering algorithms (e.g., k-means) can implement the distance metric methods only on numeric data. This paper focuses on hybrid conceptual clustering algorithm Hierarchy of Attributes and Concepts (HAC). This paper demonstrates the implementation of HAC in the diagnosis and prevention of diabetes.