PENERAPAN ALGORITMA JARINGAN SYARAF TIRUAN UNTUK PREDIKSI KELAYAKAN KREDIT DI KOPERASI BINA BERSAMA LAMPUNG UTARA

Abstract

KSP BINA BERSAMA is a financial institution which is engaged in savings and loans. This study aimed to determine melakukakan granting credit worthiness of its customers. Financial institutions have a policy lending respectively. In granting credit policies set the standard for accepting or rejecting credit risk. Ratings are eligible to receive a credit to qualify Five C, the character Customer (Character), the capacity to pay off the credit (Capacity) the ability of capital owned by the customer (Capital), guarantees that customers have to bear the credit risk (Collateral) and the financial condition of the customer (Condition ). In this study, the authors analyzed the predictions of the credit worthiness of the KSP BINA BERSAMA. The author conducted analysis using 10 patterns for determining credit worthiness ie salary, job, family card, ID card, collateral, United Nations, Business License / SITU, large and long installment loans as well as the value - average. Analyses were performed by applying the algorithm Backpropagation with 9 as the input variables and the X10 value - average as the output value. in this study I use training and testing with a pattern of 9-3-1 with three different learning rate is 0,1,0,8 and 0.5 with MSE 0.000001. From the results of the training and testing with existing customer data obtained accuracy achieve the desired target of about 88% with the average test time of 0.4 seconds

Keywords :Backpropagation,learning rate, five C, Charakter, capacit, capita, collateral, Condition, MSE