Feature Extraction and Prediction of Stock Information Based on LSTM

The stock market has been closely watched by investors because of its high risk and high return characteristics. In order to get a return from the market, they need a scientific investment method. In recent years, deep learning has been widely used in various scenarios, one of which is prediction research in financial markets. In this paper, the tushare tool is used to obtain stock data, the principal component analysis method is used to reduce the dimension of the indicators to build an LSTM model, and the extracted indicators are used as model input. Through parameter tuning and model training, a prediction model is finally obtained. The results show that the model has more advantages in long-term prediction. Using this model to predict the stock market can help investors avoid risks to a certain extent.