EARLY DETECTION OF DIABETES MELLITUS USING RANDOM FOREST ALGORITHM

Abstract

Diabetes mellitus is a deadly disease. Patients with this disease often do not realize 
that they are improving their diabetes mellitus. It is necessary to do early prevention in order 
to reduce the sudden death rate of people with diabetes mellitus. In addition, during the 
COVID-19 pandemic, which increases the risk of death for people with comorbid diabetes 
mellitus. A system model for the prediction of diabetes mellitus is needed for early diagnosis 
of this disease. By using machine learning techniques using the Random Forest algorithm 
and Information Gain can be used to predict diabetes mellitus. This model has a fairly high 
level of accuracy, which is 98.27%, precision is 97.69% and recall is 98%.