Predicting the Risk of Diabetes Using Machine Learning Algorithms

Diabetes is a chronic disorder that many people suffer from. This disease develops when the blood glucose level is high. Serious side effects, such as harm to the heart, kidneys, eyes, and other organs may result from diabetes neglected treatment. Diabetes has numerous causes, including aging, obesity, inactivity, genetics, poor food and lifestyle choices. Early identification of this illness helps lessen its negative consequences. There are many traditional methods for predicting this disease, but they are expensive. Early prediction of diabetes benefits all those at risk by providing early treatment. With the advancement of healthcare technology, machine learning algorithms can analyze large amounts of data, which can help the medical sector make more accurate and timely decisions. In this paper, artificial intelligence algorithms were used to help medical professionals predict this disease, as these technologies can greatly help the medical sector by predicting the possibility of diabetes with the utmost accuracy, thus saving time for both doctors and patients. The main goal of this study focused is to employ machine learning algorithms to analyze medical data and select the best algorithm for predicting the disease by comparing the evaluation metrics of these algorithms. The Indian diabetes dataset PIMA obtained from the Irvine Machine Learning (ML) Repository in the University of California, was used. This research applied four algorithms including the decision tree algorithm (C4.5), K-Nearest Neighbor algorithm, random forest algorithm, and support vector machine algorithm was used to predict diabetes. The experiment results illustrated that the random forest algorithm did the best overall than other models in predicting disease.