Improvement in prediction rate and accuracy of diabetic diagnosis system using fuzzy logic hybrid combination

Nowadays diabetes mellitus has become the major health problem among the people of all ages. The main problem in this type of dieses is its prediction. It is found that if diabetes mellitus is detected at early stages then it can be cured. So early detection of diabetes mellitus is important. There are different techniques with the help of which early detection of diabetes mellitus is possible. In this paper combination of three different methods used for early detection of diabetes mellitus are given. These three methods are fuzzy system, neural network, case based reasoning. By using combination of all these approaches, it is found that detection of diabetes mellitus at early stages is possible. Benefit of using these systems is that accuracy of prediction rate is higher as compared to other techniques.