Analisis Default Kartu Kredit Dengan Deep Learning Untuk Mendukung Keputusan Manajemen Keuangan Digital

Abstract

The development of the digital economy requires financial institutions to optimize risk management through data-driven analysis. This study aims to analyze the factors influencing credit card default and to develop a predictive model using a Deep Learning algorithm based on an Artificial Neural Network (ANN) to support digital financial management decision-making. The data were obtained from the public “Default of Credit Card Clients” dataset (UCI/Kaggle), consisting of 30,000 observations and 23 financial variables. The results show that the model achieved an accuracy of 81.6% and an AUC value of 0.771, with high specificity but relatively low recall. These findings indicate that deep learning is effective in capturing non-linear patterns in customer payment behavior and can serve as a decision support tool for digital financial institutions in identifying credit risk and designing more adaptive default mitigation strategies.