Creditworthiness System with K-Means Model Based Z-Score
Y A Gerhana*, S C Kusuma, W B Zulfikar, D S Maylawati, M A Ramdhani

Department of Informatics, UIN Sunan Gunung Djati Bandung
yanagerhana[at]uinsgd.ac.id

Abstract

Evaluation of the creditors ability needs to be done before the debtor provides the loan. The purpose of system design in this study is to establish a creditworthiness system by implementing K-Mean based Z-Score. The method of system design that used in this research is Rational Unified Process (RUP). The case study of thos research is in the Savings and Loans Cooperative (KSP) at Senior High School 1st Cileunyi, Bandung. K-Means method based Z-Score is used by the author as a model for solving problems, wherein the author uses three clusters with four parameters that applied to the method. Based on the results of the experiments, the application can classify prospective credits into three groups based on the four parameters. Those three groups among others: "OK" as many as 37 members, "Warning" as many as 28 members and "Bad" as many as 5 members. The results of those groups are used as a credit recommendation to the KSP.

Keywords: Clustering, K-Means, Loan, Save, Z-Score

Topic: Computer Science