Chronic Kidney Disease (CKD) Classification with KNN Algorithm and ID3 Decision Tree

Abstract

Chronic Kidney Disease is a global health problem that requires diagnosis to prevent complications.
According to the Director of Non-Communicable Disease Prevention and Control of the Indonesian
Ministry of Health, in Indonesia, Chronic Kidney Failure is the 10th leading cause of death with more than
42,000 deaths per year. Chronic kidney disease is a condition in which kidney function gradually declines.
Chronic kidney disease can occur due to various factors, including hypertension, diabetes, autoimmune
diseases, kidney infections, and kidney stones that are not treated properly. A step that can be used for
prevention is to identify the disease with data mining classification. Many methods have been used to
predict chronic kidney disease, including the K-Nearest Neighbor (KNN) & ID3 Decision Tree methods. In
this study, classification was carried out using the KNN and ID3 methods by testing data with various
percentages of test data, namely 10%, 20%, 30% and 40%. After testing, the highest calculation result of
the KNN method is in the 30% percentage test data with a value of k = 3, the accuracy obtained reaches
99.16%. While in the ID3 Decision tree method, the highest accuracy value is found in the 30% percentage
of test data with an accuracy value of 98.33%.

Keywords: Chronic Kidney Disease; Classification; K-Nearest Neighbor; Decision Tree ID3