Implementasi Metode Artificial Neural Network Untuk Analisis Resiko Kredit Pada Sistem Peer To Peer Lending

Abstract

Peer to Peer (P2P) lending is an example of implementing Financial Technology (Fintech) in the form of information technology-based lending and borrowing services. This service offers flexibility where lenders and borrowers can allocate and get capital or funds almost from and to anyone, in any amount of value, effectively and transparently, and with competitive returns. The problem with the Peer to Peer lending application is how to provide recommendations to users, especially to lenders regarding the risk of borrowing from borrowers. To solve this problem, the Artificial Neural Network method is applied. The result of this research is a web- based Peer to Peer Lending system application. The software is able to provide information on the results of recommendations to lenders regarding the type of loan. From the test results, information is obtained that the Artificial Neural Network method is suitable to be applied in a peer-to-peer lending system, with a validity level of 80%. In addition, the software is also able to provide highly reliable recommendation information to lenders regarding the type of loan, with a sensitivity level of 100% loan risk prediction, 60% positive predictive value and 100% negative predictive value.