基于EMD-LSTM模型的股指收盘价预测  

Prediction of Closing Price of Stock Index Based on EMD-LSTM Model

作　　者：刘铭 单玉莹 LIU Ming;SHAN Yuying(College 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,长春130012

出　　处：《重庆理工大学学报（自然科学）》2021年第12期269-276,共8页Journal of Chongqing University of Technology：Natural Science

基　　金：国家自然科学基金项目(61503150);吉林省自然科学基金项目(20200201157JC);吉林省教育厅科学技术项目(JJKH20191295KJ)。

摘　　要：股指是投资者用来规避股市风险的工具,为了对金融股指进行有效预测,采用了一种基于经验模态分解(EMD)和长短期记忆网络(LSTM)模型的组合预测方法,对股指进行统计性描述,发现中国3个股指的波动具有明显区别,就这一特征对数据进行建模。建立传统时间序列模型及机器学习模型共7种模型,经对比研究发现:EMD-LSTM模型在预测沪深300股指收盘价和深证成指收盘价上具有较好的效果,预测上证指数收盘价时LSTM模型具有较好的效果,从而分析出数据波动大小对于模型的预测效果有一定的影响,可以根据数据波动性来选择适合的股指预测模型。An equity index is a tool used by investors to hedge against the risks of the stock market,in order to effectively predict financial stock indices,a combined prediction method based on empirical modal decomposition(EMD)and Long Short Term Memory(LSTM)model is adopted.Firstly,the stock indices are statistically described,and the fluctuations of the three Chinese stock indices are found to be significantly different.To model the data on this feature,this paper establishes a total of seven models,both traditional time series models and machine learning models,and the comparative study finds that the EMD-LSTM model has a good effect in predicting the closing price of CSI 300 Index;the LSTM model has a good effect in predicting the closing price of Shanghai Composite Index;and the EMD-LSTM model has a good effect in predicting the closing price of SZSE Component Index.It is analyzed that the size of the data volatility has an impact on the prediction effect of the model,and that a suitable stock index prediction model can be selected based on the data volatility.

关 键 词：股票指数 金融预测 EMD-LSTM 深度学习 经验模态分解 

分 类 号：F832.5[经济管理—金融学]