摘要
针对传统企业经济危机风险预测模型精度低,财务指标变量间相关性重叠度高的缺陷,研究提出利用KPCA算法对财务指标进行降维,利用SMOTE算法实现样本数据的平衡性修正,最后结合XGBoost模型构建企业经济危机风险预测模型。实验结果显示,研究构建模型的预测精度超过90%,在4种模型当中,模型1、模型2、模型3和模型4的AUC值分别为0.902,0.863,0.771和0.694。研究结合KPCA和XGBoost构建的企业经济危机风险预测模型具有较高的精度,能够有效预测企业经济危机,为企业的发展提供保障。
In view of the shortcomings of the traditional enterprise economic crisis risk prediction model,such as low accuracy and high correlation overlap between financial index variables,KPCA algorithm is proposed to reduce the dimension of financial indicators.The SMOTE algorithm is used to correct the balance of the sample data.Finally,combined with XGBoost model,the enterprise economic crisis risk prediction model is constructed.The experimental results show that the prediction accuracy of the research model exceeds 90%.Among the four models,the AUC values of Model 1,Model 2,Model 3 and Model 4 are 0.902,0.863,0.771 and 0.694 respectively.The research combined KPCA and XGBoost to build an enterprise economic crisis risk prediction model with high accuracy,which can effectively predict the enterprise economic crisis and provide guarantee for the development of enterprises.
作者
贾宁
JIA Ning(Anhui Lvhai Vocational College of Business,Hefei 230601,China)
出处
《佳木斯大学学报（自然科学版）》
CAS
2023年第2期177-180,共4页
Journal of Jiamusi University：Natural Science Edition
