基于改进的随机森林—模糊积分模型的家庭农牧场信用评级研究  

Research on Credit Rating of Family Farms and Ranches Based on an Improved Random Forest-Fuzzy Integral Model

作　　者：李战江[1] 廖湘黔 Li Zhanjiang;Liao Xiangqian(College of Economics and Management,Inner Mongolia Agricultural University,Hohhot,010010,Inner Mongolia,China)

机构地区：[1]内蒙古农业大学经济管理学院,内蒙古呼和浩特010010

出　　处：《征信》2024年第9期53-62,92,共11页Credit Reference

基　　金：国家自然科学基金资助项目(72161033)。

摘　　要：缓解融资难融资贵问题是实现家庭农牧场可持续发展的关键,而构建家庭农牧场信用评级体系是解决该难题的基础。针对家庭农牧场信用风险评估中样本数据指标多、噪声复杂、非线性乃至高维度等问题,以内蒙古自治区12个盟市500个家庭农牧场为信用评价对象,将XGBoost算法、随机森林算法、模糊积分模型进行组合,构建了改进的随机森林—模糊积分模型,对家庭农牧场进行信用评级。分析结果表明:运用改进的随机森林—模糊积分模型对家庭农牧场进行信用评级的准确率、精确率、召回率、F1分数更高。Alleviating the problem of financing difficulty and high cost is the key to realizing the sustainable development of family farms and ranches,and constructing a credit rating system for family farms and ranches is the basis to solving the problem.Aiming at the problems of multiple sample data indicators,complex noise,non-linearity and even high-dimensionality in the credit risk assessment of family farms and ranches,500 family farms and ranches in 12 cities in Inner Mongolia Autonomous Region are taken as the credit evaluation objects.The XGBoost algorithm,random forest algorithm and fuzzy integral model are combined to construct an improved random forest-fuzzy integral model,which is used to provide credit ratings to family farms and ranches.The analysis results show that the accuracy,precision,recall and F1 score of the improved random forest-fuzzy integral model are higher for the credit rating of family farms and ranches.

关 键 词：XGBoost 随机森林 模糊积分 信用评级 家庭农牧场 

分 类 号：F832.4[经济管理—金融学] F324.1