Credit Risk Evaluation of Supply Chain Finance Based on K-Means-SVM Model

Abstract

With the rapid development of supply chain finance, it is important to evaluate its credit risk effectively. The Support Vector Machine (SVM) is designed to construct the credit risk measurement model of supply chain finance. Considering the characteristics of SVM model, we select the clustering center based on K-Means clustering algorithm and the edge points far from the clustering center as training samples to train the SVM model. Experimental results show that compared with single SVM model, the overall classification accuracy of K-means-SVM model is increased by 7.2%, and the first type error rate is reduced by 5.0%, which verifies the superiority and effectiveness of k-means-SVM model applied to enterprise credit risk assessment under supply chain finance mode.