Application of Association Rule Mining in Risk Analysis for Diabetes Mellitus

Abstract: Diabetes is a growing epidemic and non communicable disease that affects a major portion of the population in the developing countries. Prevention and overall clinical management of patients with elevated risk of developing diabetes mellitus can be aided greatly by early detection of risk of getting diabetes among patients. This is done by using Association Rule Mining of the data present in the medical records of patients. Association rule is used to discover sets of risk factors for patients at particularly high risk of developing diabetes, which are called item sets of all the interacting conditions. Association rule mining generates a very large set of rules which needs to be summarized for easy clinical use. A comparative evaluation is performed to provide guidance regarding their applicability, strengths and weaknesses.

Keywords: diabetes mellitus, association rule mining, apriori algorithm, discretization.