PEMBENTUKAN POHON KLASIFIKASI BINER DENGAN ALGORITMA CART (CLASSIFICATION AND REGRESSION TREES) (STUDI KASUS KREDIT MACET DI PD. BPR-BKK PURWOKERTO UTARA)

Abstract

Modernization and globalization of the world today has entered into various lines of Indonesian society. One consequence is people's lifestyles are more consumptive. This lifestyle causes people take out a loan at a bank or other financial institution to fulfill his wish. Some people pay the loan on credit. But in implementation, there is a variety of things causes the credit not running properly or called with problrm loan. As a service provider of credit institutions, PD. BPR-BKK Purwokerto Utara is also not free from this problem. Therefore, it is necessary to classify customers based on demographic variables using Classification and Regression Trees (CART) to minimize the chances of problem loans. Based on analysis of customer credit status data PD. BPR-BKK Purwokerto Utara, optimal classification tree formed by the number of terminal nodes as much as 6 nodes. This means there are 6 characteristics of PD customers. BPR-BKK Purwokerto Utara. And the level of accuracy of the classification tree in classifying the credit status of customers is 81.0 % . Keywords: Modernization, Globalization, Credit, Bad Credit, Customer, CART, Classification Tree .