Diabetes mellitus is a deadly disease that affects people all over the globe. An early prediction of diabetes is very beneficial as it can be controlled before the onset of the disease. Various data mining classification techniques have proven fruitful in the early detection and prediction of multiple diseases like heart attack, depression, kidney-related diseases, and many more. This paper discusses and compares various data mining techniques for the prediction of Diabetes Mellitus. Also, three widely used data mining techniques via Artificial Neural Networks (ANN), K-nearest neighbor (KNN), and Support Vector Machine (SVM) have been implemented in Matlab and the results are compared based on accuracy, recall, true negative rate, and precision.