Research on the Forecast of Stock Price Index Based on BiLSTM-GRU

The changes of the stock market are closely watched by investors and consumers. How to effectively predict the SPI will, to a large extent, enable investors to obtain greater benefits. To realize the forecast of SPI, a forecast pattern based on BiLSTM-GRU is proposed. Firstly, by collecting the relevant data of the stock market, we propose the forecast index of the SPI, and use principal component analysis and hierarchical clustering to process the data in turn. Secondly, the BiLSTM-GRU is proposed based on the data obtained after preprocessing, and the effectiveness of the pattern is verified through simulation tests. The results show that the maximum training error value of the proposed BiLSTM-GRU is only 0.029, and the forecast analysis shows that there is no significant difference between the predicted value and the actual value. The above results show that the BiLSTM-GRU pattern proposed in the study is effective in the forecast of SPI, and the feasibility of the pattern is verified in comparative experiments, providing an effective idea for stock price forecast.