Abstract

Diabetes is a serious condition that is primarily brought on by aging, high blood pressure, pregnancy, insulin, body mass index, and the diabetes gene. Remember that if diabetes is not treated, it can cause damage to other organs and lead to major health issues like high blood pressure, kidney, eye, and heart damage. For the best possible care, diabetes must be detected early. Our research uses a range of machine learning techniques to improve the accuracy of early diabetes identification in patients or persons. To anticipate when diabetes may manifest, we would employ machine learning, ensemble classification algorithms, and a dataset. Naïve Bayesian network, logistic regression (LR), K-Nearest Neighbor, random forests (RF), and support vector machines (SVM). The research results in a model that clearly demonstrates its ability to predict diabetes. Our results show that the Random Forest model performs more accurately than other machine learning techniques.