A Survey on Data Mining Methods and Techniques for Diabetes Mellitus

Detection of knowledge patterns in clinical data through data mining. Data mining algorithms can be trained from past examples in clinical data and model the frequent times non-linear relationships between the independent and dependent variables. The consequential model represents formal knowledge, which can often make available a good analytic judgment. Classification is the generally used technique in medical data mining. This paper presents results comparison of five supervised data mining algorithms. We evaluate the performance for C4.5, SVM, k-NN, Naïve Bayes and Apriori then Comparison a performance of data mining algorithms based on accuracy. The study describes algorithmic discussion of the dataset for the disease acquired from on line repository of large datasets.