Research on the Dynamic Model Construction and Intelligent Decision-making Application of Credit Scoring for Anhui SMEs from the Perspective of Digital Supply Chain

Abstract

This study focuses on the financing challenges faced by small and medium-sized enterprises in the digital economy era, innovatively constructing a dynamic credit assessment system based on multi-source government data. Taking Anhui Province as the research target, the study integrates market supervision, taxation, electricity, and meteorological data to develop a hybrid assessment model that combines the XGBoost algorithm with an industry prosperity correction mechanism. Empirical evidence demonstrates that the model significantly improves assessment efficiency (AUC 0.85), shortens approval cycles by 26.7%, reduces agricultural default rates by 19.3%, and boosts the loan approval rate of the Wuhu auto parts industry cluster by 32%. The study also proposes policy recommendations, such as establishing a provincial credit data hub and developing climate-sensitive financial products, providing a replicable "Anhui model" for the digital transformation of county-level finance.