Early Determination of Diabetes Mellitus Disease Prediction with Decision Tree Boosting

Diabetes Mellitus is a disease characterized by hyperglycemia and impaired carbohydrate metabolism with reduced performance of insulin secretion. need to know in 2017 cases of diabetics amounted to 451 million and are expected to continue to increase to reach 693 million cases until 2045. Seeing the number of cases of diabetes mellitus in the world can be done early prevention efforts to prevent the occurrence of an increase in people with diabetes mellitus with early predictions of diabetes determination. This paper implemented Adaboost to optimize decision tree performance, resulting in a better level of accuracy in decision trees and then compared to other classification methods as a comparison. A workflow system designed to create a model that can analyze diabetes mellitus using boosting methods on its application. test with six algorithms as a comparison to evaluate the results of the analyzed model. The results of the experiment showed an increase in the accuracy rate in DTBoost by 98.04 % higher than the previous Decision Tree (DT) accuracy level.