A hybrid clustering data mining technique (HCDMT) for predicting SLE

Abstract

SLE is an auto immune and complex disease. Predicting Systemic Lupus Erythematosus (SLE) is significantly challenging due to its high level of heterogeneity in symptoms. There is a limitation on the tools used for predicting SLE accurately. This paper proposes a machine learning approach to predict the disease from SLE data set and classify patients in whom the disease is active. The data purified and selected for classification improves the accuracy of the proposed method called HCDMT (Hybrid Clustering Data Mining Technique), an amalgamation of CART and k-Means, was evaluated on SLE data. It was found to predict above 95% of SLE cases.