摘要
随着移动互联网的飞速发展,基站的流量负荷问题受到普遍关注。笔者基于长短期记忆网络(Long Short-Term Memory,LSTM)模型融合自回归移动平均模型(Autoregressive Integrated Moving Average Model,ARIMA)建立LSTM-ARIMA模型,首先使用前60%训练集对LSTM进行训练,然后进行误差计算得到误差序列。接下来对所得到的误差序列使用ARIMA建模,得到测试集的预测误差,最后综合LSTM与ARIMA模型得出结果。经过实验得出模型测试集的精度为88.06%。
With the rapid development of mobile Internet,the traffic load problem of base stations has attracted widespread attention.In this paper,the LSTM-ARIMA model was established based on the LSTM model and the ARIMA model.First,the first 60%training set was used to train the LSTM,and then the 40%prediction result of the LSTM was compared with the true value to calculate the error sequence.Next,use ARIMA to model the obtained error sequence to obtain the prediction error of the test set.Finally,the results were obtained by combining LSTM and ARIMA models.After experiments,the accuracy of the model test set was 88.06%.
作者
张阳
张蕗怡
ZHANG Yang;ZHANG Luyi(School of Computer Science and Information Engineering,Hubei University,Wuhan Hubei 430205,China)
出处
《信息与电脑》
2021年第7期100-102,共3页
Information & Computer
