Comparative Analysis of Diabetes Prediction Using Machine Learning

Nowadays, one of the most common chronic diseases is diabetes mellitus, and sometimes, it may lead to death. If we predict diabetes in an early stage, then it will be helpful to take preventive measures, and it can helpful to prevent progression of the disease. In today's lifestyle, food containing large number of sugars, carbohydrates and fats which increases the high risk of diabetes. The computer-based detection helps the doctors to diagnose the disease in early stage. There are many machine learning algorithms were used for the prediction and classification of diabetes. In this paper, performance of different machine learning algorithms like support vector machine, artificial neural network, deep neural network, convolutional neural network has been compared against the parameters sensitivity, specificity and accuracy. The results were compared and tabulated to show which algorithm produces more efficient and accurate results.