The stock price forecast is a hot issue that investors are concerned about.Using artificial intelligence to forecast the stock price is a very popular forecasting method in recent years.This paper uses Neural Network to study the prediction of stock price.The main work of this paper includes the following aspects:(1)Using the traditional BP Neural Network and real stock data to forecast the price of stock.The simulation experiment method is used to analyze the applicability and limitations of BP Neural Network in stock price forecasting;(2)The traditional BP Neural Network has the disadvantages of random initialization of weights and thresholds,which affects the accuracy of the Neural Network.Therefore,this article provides a method that uses Differential Evolution algorithm to optimize the initial weights and thresholds of BP Neural Networks.A BP Neural Network stock price based on Self-adaptive Differential Evolution algorithm is established.The forecast model uses real stock data to verify the superiority of the model through simulation experiments.(3)In order to overcome the limitations of the single Neural Network prediction model,this paper also presents a Combined Neural Network stock price prediction model based on Self-adaptive Differential Evolution algorithm.The model uses BP +BP Neural Network combination,which uses two BP Neural Network with different learning rates to establish a stock price forecasting model.Finally,simulation experiments were conducted to analyze and compare the validity and superiority of the forecast model based on Combined Neural Network in stock price forecasting.Finally,the research content of this article is summarized and the follow-up work is planned. |