基于LSTM神经网络的股价短期预测模型

Stock price prediction is the most challenging problem in the field of time series forecasting. The accurate prediction of stock price can help investors to reduce the risk and improve their earnings. In this paper, the application of LSTM neural network for stock index forecasting analysis, according to market maturity and select 3 kinds of index as the research object, and combined with the stock of technical indicators, then construct several useful indictors based on the OHLC index and the technical indexes; then the data set is divided into training set and sliding window prediction; finally, compared with RNN. The experimental results show that the LSTM neural network can effectively predict the stock index price and tracking index, and compared with the prediction result of RNN, the LSTM neural network has significantly improved the prediction performance.