Financial distress prediction model of dual constraint LS-SVM based on neighborhood rough set index optimization
Guanhua Zhao1, Tang lu2, Lin Gao3
COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(12C) 56-62
1Accounting Institute of Shandong Finance University, Shandong Accounting Science Research Center, Ji'nan, China
2Accounting Institute of Shandong Finance University, Ji'nan, China
3Qingdao Vocational and Technical College of Hotel Management, Qingdao, China
In order to improve the accuracy of financial distress prediction and the effect of model forecast, and make the neighborhood rough set and genetic algorithm apply to the dual constrained least squares support vector machine. This study proposes a dual constrained least squares support vector machine prediction model based on the neighborhood rough set attribute reduction. At the same time, this study gives the steps to improve this model. The empirical results show that, after the pre-treatment of neighborhood rough set index and optimization of parameters of genetic algorithm. It not only improves the model prediction accuracy, but also reduces the model run time, therefore it confirmed that the application of the improved model to the financial distress prediction is effective.