LSTM-based Stock Prediction Modeling and Analysis

Abstract

The stock market plays an important role in the economy of a country in terms of spending and investment. Predicting stock prices has been a difficult task for many researchers and analysts. Research in recent years has shown that Long Short-Term Memory (LSTM) network models perform well in stock price prediction, and it is considered one of the most precise prediction techniques, especially when it is applied to longer prediction ranges. In this paper, we set the prediction range of the LSTM network model to 1 to 10 days, push the data into the built LSTM network model after pre-processing operations such as normalization of data, and set the optimal values of epochs, batch_size, dropout, optimizer and other parameters through training and testing. By comparing with Linear Regression, eXtreme gradient boosting (XGBoost), Last Value and Moving Average, the results show that the LSTM network model does not perform better than other models when applied to a short forecasting horizon.