基于EMD和GARCH模型的股票收盘价格预测  

Stock closing price prediction based on EMD and GARCH models

作　　者：马育欣 王纯杰 秦喜文 董小刚 MA Yuxin;WANG Chunjie;QIN Xiwen;DONG Xiaogang(School of Mathematics and Statistics,Changchun University of Technology,Changchun 130012,China)

机构地区：[1]长春工业大学数学与统计学院,吉林长春130012

出　　处：《长春工业大学学报》2021年第1期7-10,共4页Journal of Changchun University of Technology

基　　金：国家自然科学基金资助项目(11571051)。

摘　　要：首先对股票收盘价序列进行经验模式分解(EMD),对分解后的本征模函数(IMF)与残差序列分别拟合ARMA-GARCH模型。将所建模型的预测结果相加,往后预测5 d的收盘价格,并与原始序列所建ARIMA-GARCH模型的预测结果进行比较。比较结果可以帮助股票投资者预测股票市场行情。The stock closing price sequence is decomposed with the Empirical Mode Decomposition(EMD).Bothe the decomposed Intrinsic Mode Function(IMF)and residual sequence are fixed to ARMA-GARCH model.The prediction results of the model are put together first,then the closing prices in the following 5 days are predicted and compared with the output of the ARMA-GARCH model.Our prediction results may offer some references for the stock investor decision-making.

关 键 词：非平稳序列 经验模态分解 ARMA-GARCH模型 短期预测 

分 类 号：O213[理学—概率论与数理统计]