C4.5 and ID3 Comparison to Classify Credit Issue on Indonesian National Health Insurance

Indonesian citizens are required to become participants of the national health insurance in accordance with applicable laws. Some time ago, namely in 2014, the government implemented an independent system for participants who were classified as non-workers. However, it causes many Indonesian citizens to ignore the payment. This causes various financial instability from related parties such as companies and various blood health facilities to hospitals as the final reference for the guarantee claim process. In this study, we summarize and categorize insured data from mandatory health insurance and make predictions about insured delays and smooth payments. Classification is a data mining technique that classifies data based on the attachment of the data to the sample data. The classification process begins by preprocessing the data, eliminating missing values, and selecting features in the dataset so that it can classify certain data into certain classes. In the evaluation phase, the accuracy rate of C4.5 is higher than the ID3 algorithm, namely 89.7% and 88.48%. C4.5 has an efficient and simple process performance, while ID3 has a more complex process performance.