top
Articles
  • SVM Application In Stock Trendency Forecasting Based On PSO Parameter Optimization   [MASS 2012]
  • Author(s)
  • Yun LIN, Yunsheng ZHOU
  • ABSTRACT
  • In recent years, the use of support vector machine classification on the stock market forecasts is of great popular. However, forecasts using SVM to do classification needs to adjust relevant parameters (mainly the penalty parameter c and the kernel function parameter g) to achieve the ideal prediction accuracy of classification. Many researchers have tried to use cross-validation (CV, Cross Validation) ideas, to find the penalty parameter c and the kernel function parameter g. Using cross- validation proved although the idea can be obtained in some sense optimal parameters, you can avoid the occurrence of the state of over-fitting and lack-of-fitting, but the use of grid search of this non-heuristic optimization, and sometimes if you want to in a larger scale search for the best c and g within the parameters will be very time-consuming. This article addresses the classification only using the training set to find the best parameters, not only to avoid traversing all the parameters within the grid points, but also efficient and high accuracy to predict the training set and test set can be reasonably predict, making the test set Classification accuracy rate is maintained at a high level, to avoid the occurrence of the state of over-fitting and lack-of-fitting. This article applies SVM parameter optimization to predicting the Shanghai Composite Index. The experimental results show that SVM method through PSO parameter optimization in financial time series has good modeling capabilities, to achieve a better prediction.
  • KEYWORDS
  • Financial time series; SVM; Cross-Validation; Heuristic algorithm; PSO
  • References
  • [1]
    Securities Association of China. Securities Investment Analysis [M]. China Financial & Economic Publishing House [M] 2009.5:244-248.
    [2]
    HUANG Jing. Neural Network Models in Stock Prediction [J]. Artificial Intelligence and identification technologies,2007,(3).
    [3]
    Francis E.H. Tay , CAO Li-juan, Application of support vector machines in financial time series forecasting[J], The International Journal of Management Science, Omega 29 (2001) 309–317,
    [4]
    ZHANG Qing,LIU Bing-jie, SVM parameters rapid optimization method based on PSO and group training[J]. Science Technology and Engineering,2008,8(16):4613-4616
    [5]
    Vapnik V. Statistical Learning Theory[M]. Miley, New York, NY,1998
    [6]
    Lin C J . Formulations of support vector machines: a note from an optimization point of view [J]. Neural Computation 2001,13(2) 307- 317
    [7]
    LIANG Jun,Chen Can,Modified particle swarm optimization algorithm[J]. Computer Engineering and Design,2008,29(11):2893- 2896
    [8]
    Chenn-Jung Huang, Dian-Xiu Yang, Yi-Ta Chuang.Application of wrapper approach and composite classifier to the stock trend prediction [J]. Expert Systems with Applications 34 (2008) 2870–2878
    [9]
    ZHU Yu-Lian,The feature extraction and classification based on matrix models and vector subschema[D].Nanjing: Nanjing University of Aeronautics and Astronautics, 2003.

Engineering Information Institute is the member of/source content provider to

http://www.scirp.org http://www.hanspub.org/ http://www.crossref.org/index.html http://www.oalib.com/ http://www.ebscohost.com/ http://www.proquest.co.uk/en-UK/aboutus/default.shtml http://ip-science.thomsonreuters.com/cgi-bin/jrnlst/jlresults.cgi?PC=MASTER&Full=journal%20of%20Bioequivalence%20%26%20Bioavailability http://publishers.indexcopernicus.com/index.php