摘要
准确的流量预测提升通信网络技能、改善通信网络管理具有重要意义。考虑到长短期记忆(Long ShortTerm Memory,LSTM)神经网络可解决长序列数据在训练过程中的存在梯度消失问题,采用LSTM作为长期流量预测的基准模型,为提高模型准确性,在LSTM模型的基础上,引入Attention机制(Attention mechanism,AM),通过映射加权和学习参数矩阵赋予LSTM隐含状态不同的权重,减少历史信息的丢失并加强重要信息的影响作用,以完成流量的预测。
Accurate traffic prediction is of great significance to improve communication network skills and improve communication network management.Considering that long short term memory(LSTM)neural network can solve the problem of gradient disappearance of long series data in the training process,LSTM is used as the benchmark model of longterm traffic prediction.In order to improve the accuracy of the model,attention mechanism(AM)is introduced based on LSTM model to reduce the loss of historical information and enhance the influence of important information.So that different weights are given to the LSTM implicit state by mapping weighting and learning parameter matrix,so as to complete the traffic prediction.
作者
李笑雪
蒋澄杰
马册
安津石
阙佳雄
LI Xiao-xue;JIANG Cheng-jie;MA Ce;AN Jin-shi;QUE Jia-xiong(School of electronic information engineering,Shenyang Aerospace University,Shenyang 110136,China)
出处
《电脑与信息技术》
2022年第1期14-16,20,共4页
Computer and Information Technology
基金
沈阳航空航天大学大学生创新项目(项目编号:X202110143131)。
关键词
长短期记忆神经网络
注意力机制
网络流量预测
映射加权
long and short term memory neural network
attention mechanism
network traffic prediction
mapping weighting
