Diabetes Mellitus Prediction and Classifier Comparitive Study

Diabetes is a major concern of every household these days. It is a major health issue occurring at any age these days even to infants. With the help of supervised machine learning algorithm this disease can be predicted easily. Machine learning algorithms like Support Vector Machines, K Nearest Neighbor, Naive Bayes, Logistic Regression, Adaboost etc. are the classification algorithms used as a model to predict the disease. Then, we will compare the results of these algorithms, so that best out of these can be found out. Since the dataset is small in size so the probability of overfitting is more, to rescue our model to overfit, we are using those algorithms which work on small datasets (i.e. Naïve Bayes). The most efficient algorithm is used to predict diabetes on the test data or for user input.