基于XGboost模型的农业供应链金融信用风险测度研究

随着近年来农业供应链金融的发展，如何测度和控制农业供应链金融企业的信用风险变得愈加重要。选取76家农业上市企业为研究样本，选取了企业基本情况、盈利能力、营运能力、资金周转能力4个一级指标以及14个二级指标构建农业行业上市公司信用风险评估指标体系，对比分析XGboost模型和Logistic模型的信用风险评估结果。实践表明，两个模型都具有良好的预测能力，XGboost模型在性能和预测精度上略优于logistic模型。

Research on Credit Risk Measurement of Agricultural Supply Chain Finance Based on XGboost Model

With the development of agricultural supply chain finance in recent years, how to measure and control the credit risk of agricultural supply chain finance enterprises has become more and more important. In this paper, 76 listed agricultural enterprises were selected as research samples, and the credit risk assessment index system of listed agricultural companies in the agricultural industry was constructed by constructing four first-level indicators, including basic situation of enterprises, profitability, operating capacity and capital turnover capacity, and 14 second-level indicators. The credit risk assessment results of XGboost model and Logistic model were compared and analyzed. Practice shows that both models have good prediction ability, and XGboost model is slightly superior to logistic model in performance and prediction.