Diabetes Prediction using Machine Learning Classification Algorithms

Diabetes is one among the chronic diseases or metabolic diseases in which a person's blood glucose levels in the body gets increased. During this phase, the body cells will not respond properly to the insulin present in the body. Diabetes leads to high blood sugar and it is also considered as one of the deadliest diseases across the globe. Diabetes will also result in many problems if left untreated and undiagnosed. Hence, it has to be taken utmost care and a high level of accuracy is required in the diagnostic phase. With the development of the machine learning system, the researchers have gained the flexibility to predict the glucose level with utmost accuracy. In the existing system, various machine learning algorithms such as Support Vector Machine SVM, Naive Bayes NB and Random Forest RF have been separately used to predict the blood sugar and they have also achieved a prediction accuracy of up to 75%. In the proposed system, feature extraction has been included and a comparative analysis has been done with support vector machine, naive bayes and random forest algorithms. Then, the performance of the three algorithms is evaluated in various measures such as accuracy, precision, F-measure and recall.