摘要
随着投资市场的不断发展壮大以及人工智能时代的来临,通过传统的投资方法分析投资标及对投资标的未来走势进行预测已不能满足基金经理们的投资需求.本文结合深度学习中的长短时记忆神经网络(LSTM),提出一种基于注意力机制的LSTM模型对股票价格进行预测.通过对银行板块中的股票进行研究与预测,找到最优的时间步长为10,并通过对比LSTM模型和LSTM-ATT模型,验证了LSTM-ATT模型具有更高的预测准确率.
With the continuous development of the investment market and the advent of the era of artificial intelligence,using the traditional investment methods of analyzing investment targets and predicting the future trend of investment targets can no longer meet the investment needs of fund managers.Long and short memory neural network(LSTM)in deep learning has certain advantages in dealing with long time series problems,so an LSTM model based on attention mechanism is proposed to predict the stock prices.Through the research and prediction of the stocks in the banking sector,the optimal time step is found to be 10,and by comparing the LSTM model with the LSTM-ATT model,it is verified that the LSTM-ATT model has higher prediction accuracy.
作者
陶永康
张广强
李鹏
TAO Yong-kang;ZHANG Guang-qiang;LI Peng(School of Mathematics and Statistics,North China University of Water Resources and Electric Power,Zhengzhou 450046,China)
出处
《兰州文理学院学报（自然科学版）》
2023年第2期49-54,共6页
Journal of Lanzhou University of Arts and Science(Natural Sciences)
