Bankruptcy prediction using SVM models with a new approach to combine features selection and parameter optimisation

Due to the economic significance of bankruptcy prediction of companies for financial institutions, investors and governments, many quantitative methods have been used to develop effective prediction models. Support vector machines (SVM), a powerful classification method, has been used for this task, however, the performance of SVM is sensitive to model form, parameters setting and features selection. In this study, a new approach based on direct search and features ranking technology is proposed to optimize features selection and parameters setting for 1-norm and least square SVM models for bankruptcy prediction. This approach is also compared to the SVM models with parameters optimization and features selection by the popular Genetic Algorithm (GA) technique. The experimental results on a data set with 2010 instances show that the proposed models are good alternatives for bankruptcy prediction.