Credit Risk Assessment Model of Technology-based SMEs With Heterogeneous Ensemble Learning

In recent years, with the rapid development of technology-based SMEs, their credit business has increasingly attracted the attention of commercial banks. How to promote the credit business of technology-based SMEs while strengthening risk prevention is an urgent problem to be solved, but the current risk assessment models have some shortcomings. To solve this problem, this paper proposes a credit risk assessment model for SMEs based on heterogeneous ensemble learning algorithm. Firstly, this paper constructs a risk assessment system based on multi feature analysis, establishes a nebulous integrated appraisal model of credit risk based on non-financial indicators, and obtains the credit risk rating of enterprises. On this basis, BP neural network, support vector machine and random forest algorithm are used to build three enterprise default risk prediction models respectively, and then three classifiers are integrated based on stacking strategy to construct a SME default risk identification model integrating financial indicators and non-financial indicators. The model can better identify the default risk of SMEs and has strong adaptability, So as to achieve the purpose of preventing credit risk.