基于自注意力的LSTM-CNN股票走势预测模型的研究
首发时间:2021-02-24
摘要:随着我国经济的高速发展,我国的股票市场已经成为全球最受瞩目的股市之一,吸引着海内外众多投资者的目光。随着深度学习的兴起,其已经广泛地应用在生活中的各个领域,股票走势预测问题亦是众多研究人员关注的重点,使用传统的统计学方法和机器学习方法不足以应对股票这种复杂且多变的数据内容,无法挖掘出数据背后更深层次的逻辑,从而影响股票预测的准确度。本文提出了一个基于自注意力的LSTM-CNN股票走势预测模型,同时在神经网络结构和训练优化方法等方面进行改进,通过实验分析并验证了本文提出的基于自注意力的LSTM-CNN股票走势预测模型具有良好的泛化效果和普适性,能够更加准确的预测股票未来走势。
关键词: 深度学习 股票走势预测 LSTM CNN 自注意力机制
For information in English, please click here
Research on LSTM-CNN stock trend prediction model based on self-attention
Abstract:With the rapid development ofthe China economy,the China stock market has become one of the most importment stock markets in the world, attracting many investors from homeland and abroad. With the rise of deep learning, it has been widely used in various fields. The problem of stock trend prediction is also the key point for many researchers. Traditional statistical methods and machine learning methods couldn\'t deal with the complex and complicated stocks data because changeable stock datacannot be digged out the deeper logic behind the data, and affects the accuracy of stock forecasts. This paper proposes a self-attention-based LSTM-CNN stock trend prediction model, improving the neural network structure and training optimization methods. Through the experimental analysis, it has verified the self-attention-based LSTM-CNN stock trend prediction model has better effect of prediction, and it could predict the future trend of stocks more accurately.
Keywords: deep learning stock trend prediction self-attention mechanism LSTM CNN
基金:
引用
No.****
同行评议
勘误表
基于自注意力的LSTM-CNN股票走势预测模型的研究
评论
全部评论0/1000