Komparasi Metode Evaluasi Pada Credit Scoring Data Mining

Abstract

Credit Scoring is a procedure that exists in all loan company. This procedure will give the result whether an applicant is eligible to receive loan or not. Data mining approach has been a well known method to assist this procedure. Based on data, this approach helps to measure an applicant’s credit worthiness. This research is using Classification and Regression Tree (CART) as a part of classification task in credit scoring. As a matter a fact that CART is recognized as one of the best algorithms that can be used in data mining. The output of this classification algorithm is validate using evaluation methods which are usually used to measure the performance of classification model result. These evaluation methods that commonly used in this research area are k-fold validation method and holdout method. This experiment using CART Algorithm along with credit scoring public dataset. The experiment result shows that along with CART algorithm, using cross validation method gives a better performance in term of accuracy and error rate.