Deep Learning and Machine Learning Models for Scalable Credit Card Default Prediction on Big Data

This study investigates methods for predicting credit card defaults using machine learning techniques optimized for big data environments. Accurate predictions are vital for financial institutions to manage risk and optimize lending strategies at scale. We apply deep learning models and fine-tuned machine learning algorithms, focusing on feature representation learning and normalization to handle large datasets effectively. Several models, including Logistic Regression, Random Forest, and Neural Networks, were implemented with hyperparameter tuning and advanced feature engineering. Our results highlight the importance of model optimization and feature learning in improving credit risk prediction for big data, offering insights into the factors driving default risk and advancing predictive analytics in this domain.