Analysis and prediction of breast cancer and diabetes disease datasets using data mining classification techniques

Data mining is a technique to extracts some meaningful information from large amount of data. Nowadays In healthcare sector data mining is come out important field from all the fields for providing accurate prediction of diseases and deeper study of medical data. Authors are using different data mining techniques to identification of various diseases such as stoke, diabetes, cancer, hypothyroid and heart disease etc. This paper discussed the literature study of various data mining techniques in section two. In this paper, we used two disease dataset breast cancer and diabetes dataset from the UCI machine learning repository. We classified the five classification algorithms on WEKA Explorer and WEKA Experimenter interface. WEKA tool is a good classification tool used in this paper. Naïve bayes, SMO, REP Tree, J48 and MLP algorithms are used to classify breast cancer and diabetes dataset on WEKA interface. The performances of these five algorithms have been analyzed on breast cancer and diabetes dataset using training data testing mode. After analyzing the performances of all algorithm, found that naïve bayes gives 72.70%accuracy on breast cancer dataset and SMO gives 76.80%accuracy on diabetes dataset.