Google stock price prediction by the back propagation neural network

Abstract

Due to the influence of internal and external factors, stock prices often fluctuate greatly, which is also a big difficulty in the field of data mining and machine learning. This paper uses linear regression model, BP neural network model, and hard voting model to forecast Google stock. The results and accuracy of these three prediction models were observed respectively. Then a more accurate and practical model was selected to make a reasonable prediction of the closing price of Google stock in the following 15 days. The data set used is divided into two subsets, with some data used to fit the model and the rest used to evaluate the model. The results prove that the BP neural network model has the best prediction effect. Therefore, this model was selected for the next step of prediction, and the conclusive prediction result proved that the closing price of Google stock would decline in the following 15 days.