An Approach for Disease Data Classification Using Fuzzy Support Vector Machine

Data Mining has great scope in the field of medicine. In this article we introduced one new fuzzy approach for prediction of hepatitis disease. Many researchers have proposed the use of K-nearest neighbor (KNN) for diabetes disease prediction. Some have proposed a different approach by using K-means clustering for reprocessing and then using KNN for classification. In our approach Naive Bayes classifier is used to clean the data. Finally, the classification is done using Fuzzy SVM algorithm. Hepatitis diseases data set is used to test our method. We are able to obtain model more precise than any others available in the literature. The Fuzzy SVM approach produced better result than KNN with Fuzzy c-meansand Fuzzy KNN with Fuzzy c-means. Theintroduction of Fuzzy Support Vector Machine algorithm certainly has a positive effect on the outcome of hepatitis disease. This fuzzy SVM model led to remarkable increase in classification accuracy.