基于Lasso方法与Logistic回归的上市公司财务预警分析 The Financial Early Warning Model of Listed Companies Based on Lasso Method and Logistic Regression

一般情况下，上市公司财务数据指标越多，预警效果越好，但由于多种因素影响，财务指标过多会导致变量间具有多重共线性。本文提出一种基于Lasso方法的Logistic回归上市公司财务预警模型。首先应用Lasso法对高维数据进行变量选择，达到降低数据维度和消除变量间共线性的目的，再用Logistic回归法实现对上市公司财务状况的预警。仿真实验结果表明，本文提出的方法能够有效地消除数据的冗余性，提高预警的精确性，为企业经营者提供有效的参考意见。 Generally, the more the financial data indicators of listed companies are, the better the early warning is; However, due to a variety of factors, excessive financial indicators lead to multiple collinearity among variables. This paper presents a financial early-warning model of Logistic regression listed companies based on Lasso method. Firstly, the Lasso method is used to select variables for high-dimensional data, which can reduce the data dimension and eliminate the collinearity between variables. Then, the Logistic regression method is used to predict the financial status of listed companies. Simulation experiment shows that the method proposed in this paper can effectively eliminate the redundancy of data, improve the accuracy of early warning, and provide effective reference for enterprise operators.