Institusion
Universitas Katolik Darma Cendika
Author
Pramananditya, Benedicto Reinaldy
Subject
T Technology (General)
Datestamp
2024-03-15 02:31:11
Abstract :
The International Diabetes Federation (IDF) noted that in 2021 Indonesia was listed as the country ranked number 5 in the number of people suffering from diabetes mellitus, namely 19.47 people. Diabetes Mellitus is a disease that infects many people both domestically and abroad. Diabetes Mellitus is a disorder of insulin function in the body which usually has general symptoms, namely increased blood sugar in humans. In general, Diabetes Mellitus is grouped into two types, namely Diabetes Mellitus type 1 and Diabetes Mellitus type 2. The diagnosis or classification of patients into the Diabetes Mellitus category is carried out by an internal medicine doctor. However, many people, especially in Indonesia, have a low level of awareness and awareness of this disease. This is caused by a lack of knowledge about this disease and its risks as well as limited time or costs in consulting a doctor. Therefore, it is necessary to implement artificial intelligence which is applied to the expert system application for Diabetes Mellitus. The role of this expert system application is aimed at obtaining results in the form of diagnosis, prediction and consultation. This research applies the Random Forest algorithm as a classification algorithm. In its application, this expert system application uses a combined dataset from the Mutual Cooperation Hospital and public references with a total of 70 rows of data. This algorithm model training uses a ratio of training data and test data, namely 80:20 with an accuracy obtained of 100% and from the confusion matrix evaluation the results obtained were precision of 1.00, recall of 1.00, and f1 score of 1.00. From the results of the accuracy of the training model and algorithm evaluation using the confusion matrix, it can be said that the implementation of the Diabetes Mellitus expert system using the Random Forest algorithm is suitable and accurate.