Research on Financial Early Warning of Listed Companies Based on Lasso-logistic Model

The financial status is an important factor affecting the survival and development of enterprises. When we conduct financial early-warning model analysis, the selection of index variables and the estimation of model parameters directly affect the prediction accuracy of the early-warning model. Lasso is a variable selection method for shrinkage estimation. By constructing a penalty functions to realize variety selection and retaining the advantages of subset shrinkage; Lasso can be applied to time series, high-dimensional graphics discrimination and selection. In this paper, the Lasso method and the Logistic model are combined to construct an early-warning model reflecting the financial status within the enterprise. The experimental results show that the model can effectively select relatively important influencing factors and also has good predictive evokes.