Stock price forecasting based on BP neural network model of network public opinion

Abstract

with the development of Internet, many companies announced the stock information by various medium, and stock buyers comment that information as well as make rational investment strategies to maximize their profit. At present, many retail investors, who lack of channels to obtain real information, scarce professional knowledge of investment theory and easy to be affected by public opinion of Internet, make an irrational investment. This paper aims to combine the text analysis technique with comments on Vanke A (sz000002) from website to calculate the percentage of people with a positive view. This analysis method attempts to predict the impacts of investor sentiment on the stock market in the short term. Based on the network public opinion formation mode and technical indicators in stock market, this paper puts forward a BP neural network model. Then we improve the prediction accuracy of stock price by predicting the closing price.