Stock price forecasting based on fractional grey model with convolution and BP neural network

The stock price reflects the economic trend of a stock's corresponding industry, and the stock price forecast can provide a reference for economic departments to formulate different policies. The fractional grey model with convolution(FGMC(1, m)) can solve linear problems with memory, and it will be very meaningful to predict the linear part of stock prices. Because BP neural network has excellent nonlinear fitting ability, it can be used to predict the nonlinear part of stock prices. This paper proposes a combination model of FGMC(1, m)-BP to predict the future price of stocks. Since the stock price is affected by many factors, the principal component analysis method is used to reduce the dimensionality of high-dimensional stock data during modeling. The experimental results show that the FGMC(1, m)-BP model is more accurate than the prediction results of a single FGMC(1, m) or BP neural network, and it has better prediction performance than the commonly used ARIMA-BP prediction model.