A Financial Crisis Risk Prediction Method for Small and Medium Enterprises Based on XGBoost Model
In order to improve the financial crisis risk prediction accuracy of small and medium-sized enterprises,the XGBoost model is introduced to carry out the design and research of the financial crisis risk prediction method of small and medium-sized enterprises.Establish a financial crisis early warning index system for small and medium-sized enterprises,including first-level indicators such as profitability and operating capacity,and second-level indicators such as operating profit margin and re-turn on equity.To solve the problem of the correlation and overlap of index variables,the principal component analysis method is introduced to process the data.Using XGBoost,build the financial crisis prediction model,to realize the risk prediction.Through comparative experiments,we show that the ROC curve of the new prediction method is closer to the upper left,indicating that the prediction per-formance is significantly improved.
万方数据
维普
