基于改进VMD和RBF的股票预测研究

Research on stock prediction based on improved VMD and RBF

为解决股票价格预测问题,运用混沌理论对股票市场进行非线性分析,将互信息改进的变分模态分解与神经网络结合,提出MVMD-RBF价格预测模型。选择上证指数和沪深300每日收盘价作为研究对象进行LASSO变量筛选,相空间重构,最后进行混合模型预测,并选择BP、DNN、RBF、VMD-RBF四个模型进行对比分析。结果显示,MVMD-RBF预测效果优于其他模型,这证明MVMD-RBF模型对预测混沌的股票数据具有良好的效果。

In order to solve the problem of stock price prediction,the nonlinear analysis of the stock market is carried out by using chaos theory,and the variational mode decomposition with improved mutual information is combined with the neural network to propose the MVMD-RBF price prediction model.The daily closing price of the Shanghai Composite Index and CSI 300 were selected as the research objects for LASSO variable screening,phase space reconstruction,and finally mixed model prediction,and four models were selected for comparative analysis,and the results showed that the MVMD-RBF prediction effect was better than that of other models.This proves that the MVMD-RBF model has a good effect on predicting chaotic stock data.