Creative Approaches Towards Development of Computing and Multidisciplinary IT Solutions for Society

22Loan Credit Prediction Using Deep Learning Approach

Bui Thanh Hung^* and Luu Hoang Ngoc Trinh

Data Science Laboratory, Faculty of Information Technology, Industrial University of Ho Chi Minh City, Ho Chi Minh City, Vietnam

Abstract

Credit risk is a matter of great concern to banks and the government, especially after the COVID-19 epidemic and economic recession that led to bad debts. The problem of forecasting lending capacity was born to partially limit the rise in bad debt. The desired result of the problem is to decide whether to lend for each loan application. This study proposes the use of deep learning models to solve this problem. The experimental results on the Home Credit Default Risk dataset show that the LSTM model provides the highest accuracy.

Keywords: Loan credit prediction, deep learning, LSTM

22.1 Introduction

In recent years, credit risk prevention has received special attention from the research community and businesses. These studies focus on developing new methods and technologies to help financial institutions and banks reduce risk and optimize their operations.

Artificial intelligence can help financial institutions and banks spot complex trends and relationships between financial and economic factors, thereby making smarter predictions and decisions [1, 2]. Predicting the possibility of lending is a challenge in the financial field. The goal is to predict a customer’s likelihood of obtaining a loan based on their income, savings, credit scores, and ...