论文部分内容阅读
提出一种将Granger相关信息用于时间序列预测的方法,以解决时间序列预测过程中信息利用不完全的问题.首先,通过Granger相关性检验确定时间序列系统中的可利用信息;然后,利用神经网络将可利用信息抽取出来;最后,将抽取的可利用信息融入到时间序列的预测中.实验结果验证了所提出预测方法的有效性和稳定性.
A method of using Granger information for time series prediction is proposed to solve the problem of incomplete information utilization in time series prediction.Firstly, Granger correlation test is used to determine the available information in time series system. Then, The network extracts the available information. Finally, the extracted available information is integrated into the time series prediction, and the experimental results verify the effectiveness and stability of the proposed method.