Research on Credit Risk Warning Model for Enterprise Supply Chain Partners Integrating Ensemble Learning and GAN

In the context of global industrial shifts and heightened market volatility, credit risks among enterprise supply-chain partners exhibit rapid contagion, extensive reach, and deep concealment. This volatility is further compounded by complex procurement, logistics and cooperation networks, which can mask underlying financial instability until it disrupts production continuity. Traditional credit risk warning models are difficult to meet the demand for accurate warning due to data imbalance, insufficient feature extraction, and weak generalization ability of a single algorithm. Ensemble learning improves prediction stability through the collaborative advantage of multiple models, while Generative Adversarial Networks (GANs) can solve sample imbalance problems through data augmentation. The integration of these advanced methodologies provides the technical framework necessary to improve credit risk warning under imbalanced samples and complex supply-chain relationships. By combining synthetic minority samples with ensemble learning, the model supports more accurate risk identification and timely intervention.