Rule Based System for Better Prediction of Diabetes

A rule-based system [RBS] is used in the field of artificial intelligence to store, manipulate, and interpret the information in various ways. It is also a combination of human knowledge and machine intelligence together to get the exact information in the field of medicine. Today number of medical data's is generated based on patient information in various formats with missing values. In existing data mining techniques such as clustering and classification, the role of missing values plays vital role for the prediction of disease. By considering missing values may cause wrong prediction of disease in humans. To improve the accuracy of prediction in the medical data set, this paper proposes the Rule Based Classification (RBC) technique by considering the best classifier. The concept proposed RBC technique implements diabetic data set. The major drawback of diabetes is, its symptoms are not common in all humans and they have to undergo diabetic testing. Rule-based systems can be adapted and applied to a large kind of problems. RBC technique can be used to predict diabetes in patients by applying various steps, facts, symptoms to make suitable rules and decide the best rule related to disease. This paper also provides the comparative analysis of various classifiers pertaining to diabetic data set.