A study on stock closing price prediction based on improved CEEMDAN and VMD quadratic decomposition

In recent years, financial stock data have been characterized by non-linearity, non-stationarity and complex dynamics, making the prediction of stock prices extremely challenging. Currently, to address the problems of limited statistical confidence and prediction accuracy in existing studies, this paper proposes a stock closing price prediction model based on improved CEEMDAN and VMD quadratic decomposition. After simulation on the IBM stock dataset, the results show that the model achieves a goodness-of-fit \({{\boldsymbol{R}}}^2\) of 98.9469% on the test set. This demonstrates the robustness and efficiency of the model in stock market prediction and provides some help for future research.