EMD分解下基于SVR的股票价格集成预测
|
摘要
为实现对非平稳、非线性股票价格时间序列的高精度预测,提出经验模态分解下基于支持向量回归的股票价格集成预测方法EMD-SVRF(EMD and SVR based stock price integrated forecasting)。首先,运用经验模态分解方法获得股票对数收益率时间序列的本征模函数及趋势序列,然后,利用ε不敏感支持向量回归为各本征模函数及趋势序列分别建立预测模型,并计算各本征模函数及趋势项的预测值,最后,集成得到股票收益率序列预测值。实验表明,相对现有的EMD-Elman网络和ARMA-GARCH等主流股价预测方法,EMD-SVRF具有更小的拟合误差和预测误差,是一种高精度的股票价格预测方法。
In order to achieve high-precision prediction of non-stationary and non-linear stock price time series, an integrated stock price forecasting method EMD-SVRF based on empirical mode decomposition and support vector regression is proposed. Firstly, the empirical mode decomposition method is used to extract intrinsic mode functions and trend term from stock logarithmic return time series. Then, several prediction models are established for intrinsic modulus function and trend subsequence with ε-insensitive support vector regression applied, and the predicted values of each intrinsic mode function and trend term are calculated. Finally, these predicted values are integrated to be the predicted value of stock return. Experimental results demonstrate that EMD-SVRF is a high-precision stock price prediction method since it has smaller fitting and prediction error than other mainstream methods such as EMD-Elman network and ARMA-GARCH.
In order to achieve high-precision prediction of non-stationary and non-linear stock price time series, an integrated stock price forecasting method EMD-SVRF based on empirical mode decomposition and support vector regression is proposed. Firstly, the empirical mode decomposition method is used to extract intrinsic mode functions and trend term from stock logarithmic return time series. Then, several prediction models are established for intrinsic modulus function and trend subsequence with ε-insensitive support vector regression applied, and the predicted values of each intrinsic mode function and trend term are calculated. Finally, these predicted values are integrated to be the predicted value of stock return. Experimental results demonstrate that EMD-SVRF is a high-precision stock price prediction method since it has smaller fitting and prediction error than other mainstream methods such as EMD-Elman network and ARMA-GARCH.
引文
[1] 张超.基于误差校正的ARMA-GARCH股票价格预测[J].南京航空航天大学学报(社会科学版),2014,16(3):43-48.
[2] KIM H Y,WON C H.Forecasting the volatility of stock price index:A hybrid model integrating LSTM with multiple GARCH-type models[J].Expert Systems With Applications,2018,103:25-37.
[3] 李祥飞,张再生,黄超,等.基于EMD与STSA混合方法的金融收益信息提取与预测[J].系统工程,2014,32(2):138-144.
[4] 徐茜茜.基于支持向量机的股票价格预测及投资策略研究[D].西安:西北大学,2018.
[5] 刘海飞,李心丹.基于EMD方法的股票价格预测与实证研究[J].统计与决策,2010(23):131-134.
[6] 张东旭.基于ARMA-GARCH模型族的上证指数收益率波动的实证分析[D].北京:清华大学,2016.
[7] 于卓熙,秦璐,赵志文,等.基于主成分分析与广义回归神经网络的股票价格预测[J].统计与决策,2018,34(18):168-171.
[8] HUANG N E,SHEN Z,LONG S R,et al.The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis[J].Proceedings of the the Royal Society A:Mathematical,Physical and Engineering Sciences,1998,454:903-995.
[9] WU Z,HUANG N E.A study of the characteristics of white noise using the empirical mode decomposition method[J].Proceedings of the the Royal Society A:Mathematical,Physical and Engineering Sciences,2004,460(2046):1597-1611.
[10] DRAGOMIRETSKIY K,ZOSSO D.Variational mode Decom-position[J].IEEE Transactions on Signal Processing,2014,62(3):531-544.
[11] 李合龙,冯春娥.基于EEMD的投资者情绪与股指波动的关系研究[J].系统工程理论与实践,2014,34(10):2495-2503.
[12] NAVA N,MATTEO T D,ASTE T.Dynamic correlations at different time-scales with empirical mode decomposition [J].Physica A:Statistical Mechanics and its Applications,2018,502:534-544.
[13] 杨博楠,张智军,肖冰松,等.基于EMD阈值算法的脉冲涡流信号消噪[J].计算机工程与应用,2017,53(2):260-264.
[14] 莫赞,赵冰,黄艳莹.基于经验模态分解-自回归组合模型的网络舆情预测[J].计算机应用,2018,38(3):615-619.
[15] 谢赤,郑林林,孙柏,等.基于EMD 和Elman 网络的人民币汇率时间序列预测[J].湖南大学学报(自然科学版),2009,36(6):89-92.
[16] 李媛.基于经验模态分解方法的房地产价格重大影响因素及短期预测分析[D].大连:东北财经大学,2017.
[17] 秦宇.应用经验模态分解的上海股票市场价格趋势分解及周期性分析[J].中国管理科学,2008,16(S1):219-225.
[18] 王德青,王斐斐,朱万闯.基于EMD技术的非平稳非线性时间序列预测[J].系统工程,2014,32(5):138-143.
[19] 姚海博,茹少峰,张文明.基于动态交易量预测的VWAP算法交易卖出策略[J].运筹与管理,2015,24(2):15-220.
[2] KIM H Y,WON C H.Forecasting the volatility of stock price index:A hybrid model integrating LSTM with multiple GARCH-type models[J].Expert Systems With Applications,2018,103:25-37.
[3] 李祥飞,张再生,黄超,等.基于EMD与STSA混合方法的金融收益信息提取与预测[J].系统工程,2014,32(2):138-144.
[4] 徐茜茜.基于支持向量机的股票价格预测及投资策略研究[D].西安:西北大学,2018.
[5] 刘海飞,李心丹.基于EMD方法的股票价格预测与实证研究[J].统计与决策,2010(23):131-134.
[6] 张东旭.基于ARMA-GARCH模型族的上证指数收益率波动的实证分析[D].北京:清华大学,2016.
[7] 于卓熙,秦璐,赵志文,等.基于主成分分析与广义回归神经网络的股票价格预测[J].统计与决策,2018,34(18):168-171.
[8] HUANG N E,SHEN Z,LONG S R,et al.The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis[J].Proceedings of the the Royal Society A:Mathematical,Physical and Engineering Sciences,1998,454:903-995.
[9] WU Z,HUANG N E.A study of the characteristics of white noise using the empirical mode decomposition method[J].Proceedings of the the Royal Society A:Mathematical,Physical and Engineering Sciences,2004,460(2046):1597-1611.
[10] DRAGOMIRETSKIY K,ZOSSO D.Variational mode Decom-position[J].IEEE Transactions on Signal Processing,2014,62(3):531-544.
[11] 李合龙,冯春娥.基于EEMD的投资者情绪与股指波动的关系研究[J].系统工程理论与实践,2014,34(10):2495-2503.
[12] NAVA N,MATTEO T D,ASTE T.Dynamic correlations at different time-scales with empirical mode decomposition [J].Physica A:Statistical Mechanics and its Applications,2018,502:534-544.
[13] 杨博楠,张智军,肖冰松,等.基于EMD阈值算法的脉冲涡流信号消噪[J].计算机工程与应用,2017,53(2):260-264.
[14] 莫赞,赵冰,黄艳莹.基于经验模态分解-自回归组合模型的网络舆情预测[J].计算机应用,2018,38(3):615-619.
[15] 谢赤,郑林林,孙柏,等.基于EMD 和Elman 网络的人民币汇率时间序列预测[J].湖南大学学报(自然科学版),2009,36(6):89-92.
[16] 李媛.基于经验模态分解方法的房地产价格重大影响因素及短期预测分析[D].大连:东北财经大学,2017.
[17] 秦宇.应用经验模态分解的上海股票市场价格趋势分解及周期性分析[J].中国管理科学,2008,16(S1):219-225.
[18] 王德青,王斐斐,朱万闯.基于EMD技术的非平稳非线性时间序列预测[J].系统工程,2014,32(5):138-143.
[19] 姚海博,茹少峰,张文明.基于动态交易量预测的VWAP算法交易卖出策略[J].运筹与管理,2015,24(2):15-220.