PENERAPAN MULTIVARIATE ADAPTIVE REGRESSION SPLINES UNTUK ANALISIS FAKTOR YANG MEMPENGARUHI KELAYAKAN NASABAH YANG MENGAJUKAN PEMBIAYAAN

Abstract

Not all customers who apply for financing will be accepted by the bank. This is to avoid risks that often occur in the financing process, namely bad financing. One way to avoid this risk is to find out the factors that affect the eligibility of customers who apply for financing using the MARS method. This research was conducted at BSI Bojonegoro branch office using data on customers who applied for financing from January to March 2023, namely 75 customers. The response variables used are binary with categories of customers who do not get financing and customers who get financing. While the predictor variables used are BI checking (X1), job background (X2), type of financing (X3), number of dependents (X4), working period (X5), income (X6), plafond (X7), margin (X8) and DSR (X9). Based on the analysis, it was found that the factors had a significant influence on the eligibility of customers applying for financing were DSR which contributed 100%, income 48%, employment background 45%, margin 42%, plafond 26% and BI checking 17%. Furthermore, the MARS model obtained is used to classify eligible and unfit customers with an accuracy rate of 92%. From this research, it is expected to minimize customers who are stuck in making payments and minimize financing risks at BSI Bojonegoro branch office