Yuan Yan1, Ling Zhao2, Yutu Tang3, Tianxiao Luo4
1Graduate School, Jose Rizal University, Manila 0900, Philippines
2Business School, Changsha Commerce and Tourism College, Changsha, Hunan 410000, China
3School of Geography and tourism, Hunan University of Arts and Science, Changde, Hunan 415000, China
4Guangxi Forest Resources and Environment Monitoring Center, Nanning, Guangxi 530000, China
After more than half a century of development, there are numerous empirical analyses on financial early warning models, most of which take listed companies as the research object. The financial crisis early warning of listed companies is a small sample, and the application of neural network and logistic regression model has the problem of over-fitting, which leads to the effect of the model is not obvious. In recent years, the support vector machine method given risk minimization criteria has been widely used in financial crisis early warning, but few scholars combine SVM and penalty letter SCAD methods for financial early warning analysis. Therefore, this paper proposes an improved variable selection method Scad and support vector machine (SVM) combined model algorithm to select the best indicators and used for financial crisis early warning, and then based on the Shanghai and Shenzhen A-share listed companies as the research sample comparison Empirical effects of SVM, Lasso-SVM and Dantzig-SVM, the research results show that the selection of variables can greatly improve the accuracy of the financial crisis early warning model, and the SCAD method in the variable selection method has more advantages than Lasso and Dantzig. The SCAD-SVM model proposed has an accuracy rate of over 96% on both the training set and the test set in this paper. The model has a good classification effect and a strong economic interpretation ability. The research results of this paper not only enrich the financial early warning model, but also greatly improve the accuracy of financial crisis early warning, and provide a theoretical basis for the scientific decision-making of enterprises and shareholders.
Support vector machines; Financial crisis warning; Smoothly clipped absolute deviation
Yuan Yan, Ling Zhao, Yutu Tang, Tianxiao Luo. Research on Early Warning Model of Financial Crisis Based on SCAD-SVM. Academic Journal of Business & Management (2021) Vol. 3, Issue 5: 127-132. https://doi.org/10.25236/AJBM.2021.030520.
 Cristianini, N., & Shawetaylor, J. (2000). An Introduction to Support Vector Machines. China Machine Press.
 Fan, J., & Peng, H. (2004). On Nonconcave Penalized Likelihood With Diverging Number of Parameters. The Annals of Statistics, 32(3).
 Hearst, M. A., Dumais, S. T., Osman, E., Platt, J., & Scholkopf, B. (1998). Support vector machines. IEEE Intelligent Systems & Their Applications, 13(4), 18-28.
 Kim, K. J. (2003). Financial time series forecasting using support vector machines. Neurocomputing, 55(1/2), 307-319.
 Lawrence, S., Giles, C. L., & Tsoi, A. C. (1997). Lessons in Neural Network Training: Overfitting May be Harder than Expected. Paper presented at the Fourteenth National Conference on Artificial Intelligence & Ninth Innovative Applications of Artificial Intelligence Conference.
 Min, J. H., & Lee, Y. C. (2005). Bankruptcy prediction using support vector machine with optimal choice of kernel function parameters. Expert Systems with Applications, 28(4), 603-614.
 Vapnik, V., & Vapnik, V. (1995). The Natural of Statistical Learning Theory.
 Vapnik, V. N. (1998). Statistical Learning Theory (Adaptive and Learning Systems for Signal Processing, Communications and Control Series).
 Cui J, Guo PJ, & Xia ZM. (2011). The properties of adaptive Lasso under Poisson log-linear regression model. Journal of Northwest University (Natural Science Edition), 41.
 Zhang, J. F. (2010). Research on SCAD based on simulation. Dalian University of Technology. Xiaomin Luo, Peng, D. T. (2020). Study on SCAD regular least square regression. Journal of Guizhou University (Natural Science Edition)(4).