Business failure forecasting models and the selection of explanatory variables: The case of France
Zammel, Madiha
Business failure forecasting models and the selection of explanatory variables: The case of France - 2020.
15
The purpose of this paper is to establish different business failure prediction models and to investigate the timeline of events leading to the failure of French companies. The models used are the logistic regression and neural network models. The variables introduced in these models were selected on the basis of three techniques: principal component analysis (PCA), LASSO (the least absolute shrinkage and selection operator) and linear discriminant analysis (LDA). The results of our models show that the LASSO technique makes it possible to obtain the best “rate of good rankings.” In addition, the larger the horizon of analysis is, the smaller the forecasting capacity of the model.
Business failure forecasting models and the selection of explanatory variables: The case of France - 2020.
15
The purpose of this paper is to establish different business failure prediction models and to investigate the timeline of events leading to the failure of French companies. The models used are the logistic regression and neural network models. The variables introduced in these models were selected on the basis of three techniques: principal component analysis (PCA), LASSO (the least absolute shrinkage and selection operator) and linear discriminant analysis (LDA). The results of our models show that the LASSO technique makes it possible to obtain the best “rate of good rankings.” In addition, the larger the horizon of analysis is, the smaller the forecasting capacity of the model.
Réseaux sociaux