A Comparative Study of Machine Learning Algorithm in Diabetics Risk Prediction

Diabetes is one of the most prevalent chronic illnesses and can affect anyone, regardless of age. Many disorders strike when the glucose or sugar level is too high. Many complications brought on by diabetes contribute to a high rate of diabetic patients being readmitted. With the aid of machine learning, the correct diagnosis can be made after the system has access to sufficient, accurate, and comprehensive information about the issue. It relies on the idea of training and testing the machine with the necessary algorithm that can produce effective results for process execution. Linear regression is one of the machine learning algorithms that we compare and analysis in this research article. Our research focuses on the accuracy, performance, and algorithm criteria that are used for medical diagnosis, and we discovered that the Bayesian Classifier, Multilayer Perceptron (MLP), K-Nearest Neighbour (kNN), Random Forest (RF), and Support Vector Machine (SVM) combination can produce the best results and most effective model for diabetic prediction. By altering the pre-trained model’s layer, the outcome can also be improved.