首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于误差同步预测的SVM金融时间序列预测方法
引用本文:李祥飞,张再生.基于误差同步预测的SVM金融时间序列预测方法[J].天津大学学报(自然科学与工程技术版),2014(1):86-94.
作者姓名:李祥飞  张再生
作者单位:天津大学管理与经济学部,天津300072
基金项目:国家自然科学基金资助项目(70971097).
摘    要:在现有支持向量机(SVM)方法的基础上提出对预测误差进行同步预测的双重预测方法,利用预测到的误差对初步预测值进行校正以提高预测精度.针对误差序列非线性、非平稳以及系统动力信息不足的特点,将经验模态分解(EMD)和支持向量机(SVM)方法结合引入误差序列的预测中.对误差序列的预测分别运用初步训练误差和测试误差对预测集合的误差进行预测,将所得到的误差序列分解为若干固有模态分量(IMF),根据各个IMF不同尺度的特点,选择不同的参数对其进行预测,最终合成原始序列的误差预测值,将所预测到的误差与初步原始序列预测值结合,得到最终的预测值.仿真结果表明该方法能够很好地解决预测滞后性和拐点误差大的缺点,相对于普通的SVM预测方法具有更好的预测精度.

关 键 词:误差预测  经验模态分解  支持向量机  时间序列

Support Vector Machine Method for Financial Time Series Prediction Based on Simultaneous Error Prediction
Li Xiangfei,Zhang Zaisheng.Support Vector Machine Method for Financial Time Series Prediction Based on Simultaneous Error Prediction[J].Journal of Tianjin University(Science and Technology),2014(1):86-94.
Authors:Li Xiangfei  Zhang Zaisheng
Institution:(College of Management and Economics, Tianjin University, Tianjin 300072, China)
Abstract:A double prediction method by means of synchronous prediction of the prediction error was proposed based on the existing support vector machine (SVM) method, and the predicted error was used to correct the preliminary predicted values in order to improve the prediction accuracy. Considering that the error sequence may have features of non-stationarity, nonlinearity system and insufficient information of system dynamics, the empirical mode decomposi-tion (EMD) method was used and embed into the support vector machine method to predict the error series according to the preliminary training error and test error respectively. In order to get the final prediction error, different parame-ters were chosen to forecast the error sequence decomposed into several intrinsic mode function(IMF) components according to the different scale characteristics of each IMF. The final prediction results were obtained by using predic-tion error to correct the preliminary predicted values. Simulation results show that the new method which can solve the problems of forecast hysteresis and inflection point error effectively has better prediction precision.
Keywords:error prediction  empirical mode decomposition  support vector machine  time series
本文献已被 CNKI 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号