摘要
随着深度学习技术的发展,卷积神经网络和循环神经网络等模型在越来越多的领域表现出卓越的性能.时间序列数据的分析与预测是一个重要的研究领域,并且传统的时间序列模型在存在复杂依赖关系的应用场景的效果有待提高.鉴于此,主要探讨将循环神经网络应用于时间序列分析与预测,并与传统的Auto-Regressive(AR)模型进行比较,实验结果表明循环神经网络对时间序列的分析和预测性能有明显提高.
With the development of deep learning,convolutional neural network and recurrent neural network show excellent performance in more and more fields.Time series analysis and forecasting is a very important research field,but traditional models in this field need to be improved when dealing with complex temporal dependencies.Therefore,this paper discusses how to apply the recurrent neural network to the time series analysis and forecasting.Comparing with Auto-Regressive(AR)model,the experiments show that the recurrent neural network performs better than the traditional method.
作者
马恺
MA Kai(Department of Basic Courses,Liaoning University of International Business and Economics,Dalian,Liaoning 116052,China)
出处
《沧州师范学院学报》
2022年第1期44-48,共5页
Journal of Cangzhou Normal University
关键词
深度学习
时间序列
循环神经网络
随时间反向传播
deep learning
time series
recurrent neural network
back-propagation with time
