Early Detection of Diabetes Using Supervised Learning Approach

Diabetes has been labelled a chronic health condition with the potential to spark a global health care crisis, it is an illness which is caused by high glucose level in a human body, according to World Health Organization (WHO), the morbidity figure may reach approximately 642 million by 2040, which means one in a ten may suffer from diabetes due to unhealthy lifestyle and lack of exercise. The condition should not be overlooked since, if left untreated, it can lead to serious complications for a person's heart, kidneys, blood pressure, eyes, and other body organs. Early detection of diabetes will help in controlling it. It would provide the human life in safe way that detection of diabetes with a better accuracy. There are factors that determines the kind of diabetes mellitus in a patient which are: glucose, pregnancy, skin thickness, blood pressure, insulin, body mass index (BMI), function of the diabetes pedigree, age, and outcome. This study developed a reliable method for spotting diabetes symptoms early, with the recent development in machine learning which has a great impact on the detection and classification of diabetes. K-Nearest Neighbor (KNN), Logistic Regression (LR), Support Vector Machine (SVM), and Random Forest (RF) were the four machine learning techniques used to predict diabetes. Each of the techniques were trained and the results shows that, when compared to other approaches employed in this study, RF has the highest accuracy of 87% in accurately, the result will help the human raters in predicting effectively patients with diabetics.