DEVELOPMENT OF MACHINE LEARNING APPLICATION FOR THE EARLY DETECTION OF DIABETICS

ABSTRACT Diabetes is a chronic metabolic disease that causes high blood sugar levels. Over time, this condition causes major harm to the heart, blood vessels, eyes, kidneys and nerves. In Nigeria, diabetes frequently contributes to sudden mortality, amputations, kidney failure, heart attacks, strokes, and blindness. In order to achieve universal health coverage, it is necessary to implement all stages of prevention and control (primary, secondary, and tertiary) at once due to the rapidly escalating diabetes epidemic in Nigeria. Given the rapidly spreading diabetes epidemic in Nigeria, it is necessary to implement all primary, secondary, and tertiary prevention and control measures at the same time in order to achieve universal health coverage. For the best possible treatment to start, diabetics must be correctly identified. Usually, the diagnosis can be determined with the help of a thorough patient history and physical examination. In this study, a machine learning algorithm was created for the classification of diabetics who had received an early diagnosis. With the help of medical professionals, 400 training datasets were developed and used to train the application. A test run of the machine learning application demonstrates that it can distinguish between diabetes and non-diabetic symptoms from those entered into the system. The technology will be helpful in the early diagnosis of diabetes, according to the pertinent stake holders.