FUZZY UNORDERED RULE USING GREEDY HILL CLIMBING FEATURE SELECTION METHOD: AN APPLICATION TO DIABETES CLASSIFICATION

Diabetes classification is one of the most crucial applications healthcare diagnosis. Even though various studies have been conducted in this application, problem remains challenging. Fuzzy logic techniques recently obtained impressive achievements different application domains especially medical technique not able to deal with data a large number input variables constructing model. In research, fuzzy using greedy hill climbing feature selection methods was proposed for diabetes. A dataset 520 patients from Hospital Sylhet Bangladesh used train and evaluate classifier. Six criteria were considered authenticate results Comparative analysis proved effectiveness classifier against Naive Bayes, support vector machine, K-nearest neighbour, decision tree, multilayer perceptron neural network classifiers. Results demonstrated potential analyzing diabetes patterns all criteria.