In recent years, deep learning technology has been widely used in the financial industry, especially in the research of stock trend prediction. Aiming at the two aspects that most stock prediction algorithms do not make full use of the data and the model does not pay enough attention to the relationship between stocks, this paper establishes a graph attention network model based on technical indicators to better predict the stock trend. On the CSI 300 data set, we compare the model with the existing deep learning model and the single factor model in factor investment. A large number of experiments show that our model not only has higher prediction accuracy, but also has better profitability in practical application.