Stock Price Combination Forecast Model Based on Regression Analysis and SVM

Abstract:

According to the features of stock price change, a new forecast model, based on the regression analysis and SVM, is proposed to solve the problem of the stock price prediction. First, the regression analysis model is used to forecast stock prices, and then SVM was established to forecast and correct the error. The combined predictive values are obviously better than single method. Empirical analysis shows that the stock price based model based the regression analysis and SVM model significantly improved the forecast accuracy, it shows that the method in this paper is worth to be extended and applied.

• Periodical: Applied Mechanics and Materials (Volume 39)
• Pages: 14-18
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.39.14
• Online since: November 2010
• Authors: Feng Kong, Guo Ping Song
• Keywords: Error Correction, Regression Analysis, Stock Price Estimate, SVM Model

References

[1] Van Gestel, T, et al. Financial time series prediction using least squares support vector machines within the evidence frame. IEEE Trans. Neural Networks, Vol. 12(2001), pp.809-821.

DOI: 10.1109/72.935093

[2] LI Zong-wei, Wang Mei-yuan, Zheng Shu-hua. Stock price prediction based on Radial Basis Function neural network, University of Shanghai for Science and Technology, Vol. 24(2002), pp.81-86.

[3] A. M. Lanchlan, An improved novelty criterion for resource allocating networks, IEE Artificial Neural Networks, Conference Publication, Vol. 440(1997), pp.48-52.

DOI: 10.1049/cp:19970700

[4] D. Srinivasan , S . S. Tan, C. S. Chang and E. K. Chan. Practical implementation of a hybrid fuzzy neural network for one-dayahead load forecasting. IEE Proc. Gener. Transm., Vol. 11(1998), pp.687-692.

DOI: 10.1049/ip-gtd:19982363

[5] WANG Yu-hong, HUANG De-xian，GAO Dong-jie，etc. Nonlinear Predictive Control Based on Support Vector Machine. Information and Control, Vol. 33(2004), pp.133-136.

[6] L. J. Cao, Francis E. H. Tay. Support Vector Machine With Adaptive Parameters in Financial Time Series Forecasting. IEEE Trans. NEURAL NETWORKS, Vol. 14( 2003), pp.1506-1518.

DOI: 10.1109/tnn.2003.820556

[7] YIN hong一cai，ZHAO Cun—yan，WAN Jia—qiu. Research of Stock forecasts Based on Neural Networks. JOURNAL HARBIN UNIV．SCI．& TECH, Vol. l2(2007), pp.82-85.

[8] ZHANG Xue-jun ，CHEN Gang1, ZHOU Jie1, MAAi-jun ，etc. Medium and long-term load forecast based on PSO-RSVR, Power System Protection and Control. vol. 37(2009), pp.77-81.

[9] Jin-pei Wu, De-shan Sun, Modern Data Analysis, China Machine Press, Beijing, (2006).

[10] Tao Yang. SVM multiuser detection based on heuristic kernel, High Tecnology Letters, Beijing, June. 2007, pp.2-3.