Abstract
In the recent years, google has become one of the most powerful companies in the world, due to its big market dominance. More and more people want to predict the stock price of google, however changes in the stock price are hard to find because they combine with social and economic development. Therefore, many different models which can be divided into traditional-based model, machine learning and deep learning models are designed to improve the accuracy of stock price prediction. This paper firstly compared three high-frequency used different models based on different aspects: autoregressive integrated moving average (ARIMA) model, eXtreme Gradient Boosting (XGBOOST) model and Long short-term memory (LSTM) model. mean absolute error (MAE), mean squared error (MSE), rooted mean squared error (RMSE), r-squared(R2) are presented due to the performance of models. Empirical results show that XGboost model provide more accurate approximation than ARIMA and LSTM models. In addition, the accuracy of LSTM is the worst.