The Application Of EEMD Decomposition Method In Analysis And Forecast Of China's Stock Market

As China's economy has become more and more closely linked with the economies of the world,its mutual influence has become higher and higher.The financial market is the core content of the national economic development process.Governments and various financial investment institutions are actively studying and analyzing the laws of financial market changes.Its ultimate goal is to improve the efficiency of financial investment and manage the financial market more effectively.The stock market occupies an important position in the development of the financial market.Therefore,it is of great significance for the development of the economy to conduct in-depth research on its fluctuations.In recent years,with the rapid development of China's financial market and the growing variety of investment types,quantitative investment has gradually stepping into people's vision,which has also promoted in-depth research in financial econometric analysis.As the most common observational data in financial markets,time series is a true portrayal of market behavior.The potential laws of markets can be found by quantitative analysis,which provide theoretical basis and technical support for investment decision-making.It also facilitates follow-up risk management,asset pricing,and product design.This paper introduces the development of Ensemble Empirical Mode Decomposition method and its related theoretical basis,and expounds the principle and parameter selection of Support Vector Machine.In empirical research,this article takes the daily closing price data series of Shenzhen composite index in our country as the research object,using the empirical modal decomposition method to analyze its volatility and periodicity.And based on this,using the SVM model to predict,analyze and compare the different forecasting methods.Firstly,this paper analyzes the Shenzhen Composite Index using the set empirical mode decomposition method and decomposes the original data sequence into a finite number of intrinsic modal functions and a trend term.Then the original time series was constructed into a high frequency part,a low frequency part and a trend part based on EEMD.By analyzing the volatility and periodicity of each functions,it suggested the different volatility features of the Shenzhen Composite Index with different scales.Finally,this paper selected the data of the past two years for EEMD and Intrinsic Mode Function Reconstruction.Then we predict and analyze based on the mean ratio of the reconstruction sequence.We get the predictive value of each sequence,and get the final prediction by combining three predictive values.By comparing the prediction results of the forward step prediction and the forward five-step prediction method,the effectiveness of the five-step prediction method for the future prediction is verified.Compared with the single SVM model prediction results,it shows that using the EEMD method to process the data can effectively improve the prediction accuracy,and the application space is more extensive.