Prediksi Tingkat Risiko Kredit dengan Data Mining Menggunakan Algoritma Decision Tree C.45

Abstract

Finance companies in providing credit conduct data analysis first to reduce credit risk. When customers do not pay credit smoothly, it will harm the company. For this reason, credit analysis is an important factor to minimize financial risk. So, it takes a predictive analysis of the level of credit risk based on data or files from customers. This study aims to predict the level of credit risk with data mining using the C.45 decision tree algorithm. There are two classes of risk level predictions, namely current and non-current. The C.45 decision tree algorithm has a function to find knowledge or patterns of characteristic similarity in a particular group or class. In this study, the C.45 algorithm was implemented and analyzed using the WEKA application. From the results of the evaluation using the confusion matrix, the accuracy generated for 1,153 training data with 91 testing data and the six attributes used produces an accuracy of 79%