基于注意力机制和LSTM网络的股价预测

Stock Price Prediction Based on Attention Mechanism and Long Short-Term Memory Network

摘要: 随着全球经济金融一体化，股票市场的交易规模不断增大，传统的计量经济学难以充分学习股票市场的非线性变化。本文运用深度学习中的长短期记忆神经网络(LSTM)作为基本模型对股票数据的开盘价进行预测，实验选用广聚能源(000096)和北方国际(000065)两组数据，之后在LSTM网络模型中引入注意力机制(Attention Mechanism)。经实验该模型预测精度有明显提升，说明AM-LSTM网络模型在股票预测领域具有一定可靠性。

Abstract: With the integration of global economy and finance, the transaction scale of stock market keeps increasing, so it is difficult for traditional econometrics to fully learn the nonlinear changes of stock market. In this paper, long and short-term memory neural network (LSTM) in deep learning is used as the basic model to predict the opening price of stock data. In the experiment, data of Guangju Energy and Norinco International are selected, and then Attention Mechanism is introduced into THE LSTM network model. Through experiments, the prediction accuracy of the model is significantly improved, indicating that the AM-LSTM network model has certain reliability in the field of stock prediction.