AdaBoost Based C4.5 Accuracy Improvement on Credit Customer Classification

Abstract

Credit has become commonplace in today's society. Many people choose to take credit to support their economy, both for business capital and other activities. In order to produce the right decision, credit recipient customers can be classified according to their possible payment performance. The research was conducted to improve the accuracy of the Decision Tree C.45 algorithm by using Adaboosting in classifying credit customers, to get the most optimal accuracy in terms of credit customer classification. With AdaBoost improvement, the accuracy of the c4.5 algorithm was significantly improved from 45.38% to 100% and has a much higher accuracy rate when compared to naive bayes which has been improved as well as a comparison.