A Hybrid Device of Self Organizing Maps (SOM) and Multivariate Adaptive Regression Splines (MARS) for the Forecasting of Firms’ Bankruptcy
Javier DE ANDRES (),
Pedro LORCA and
Francisco Javier DE COS JUEZ Additional contact information Javier DE ANDRES: University of Oviedo, Department of Accounting, Spain
Fernando SÁNCHEZ-LASHERAS: University of Oviedo, Department of Construction and Manufacturing Engineering, Spain
Pedro LORCA: University of Oviedo, Department of Accounting, Spain
Francisco Javier DE COS JUEZ: University of Oviedo, Department of Exploitation and Exploration of Mines, Spain
This paper proposes a hybrid approach to the forecasting of firms’ bankruptcy of Spanish enterprises from the construction sector. Our proposal starts splitting the group of healthy companies into two subgroups: borderline and non-borderline companies. Borderline companies are healthy companies with marked financial similarities with bankrupt ones. Then, each subgroup is divided in clusters according to their financial similarities and then each cluster is replaced by a director vector which represents the companies included in the cluster. In order to do this, we use Self Organizing Maps (SOM). Once the companies in clusters have been replaced by director vectors, we estimate a classification model through Multivariate Adaptive Regression Splines (MARS). Our results show that the proposed hybrid approach is much more accurate for the identification of the companies that go bankrupt than other approaches such as a multi-layer perceptron neural network and a simple MARS model.
More articles in Journal of Accounting and Management Information Systems from Faculty of Accounting and Management Information Systems, The Bucharest University of Economic Studies Series data maintained by Cristina Geambasu ().