Classification of Diabetes Dataset with Data Mining Techniques by Using WEKA Approach

Diabetes is a very common disease and beside it causes serious health problems such as fatal kidney damage or blindness; it may lead the patient to death. There is no exact cure for this disease yet but it is manageable with medication and diet. In this manner, importance of correct diagnosis of diabetes is very important to identify the diseases in early stage and take necessary precautions. There is a lot of data accumulated on this subject, as there are so many patients with this condition. This makes it possible for researchers to use data mining techniques on this subject. This study is proposed to classify diabetes by using data mining techniques. The dataset which has been obtained from UCI machine learning depository contains 520 instances, each having 17 attributes. Seven different classification algorithm including Bayes Network, Naïve Bayes, J48, Random Tree, Random Forest, k-NN and SVM have been studied on this dataset. Obtained results indicated that k-NN performed the highest accuracy with 98.07% and this algorithm is the best method to identify and classify diabetes diseases on studies dataset.