摘要
利用该模型对2019年的铁路月度客运量数据进行预测,并与实际客运量进行对比分析,实验表明,LSTM-Prophet算法模型预测值的MAPE值为1.91%,MAE值为559.07,RMSE值为691.82,三个评价指标值都低于SARIMA模型、LSTM模型和GM(1,1)模型的评价指标值,所以LSTM-Prophet算法模型更能准确预测客运量数据,为决策提供参考价值。
The model is used to predict the monthly railway passenger volume data in 2019 and compared with the actual passenger volume.The experiment shows that the MAPE value of LSTM prophet algorithm model is 1.91%,MAE value is 559.07 and RMSE value is 691.82.The three evaluation index values are lower than those of SARIMA model, LSTM model and GM(1,1)model, Therefore, LSTM prophet algorithm model can more accurately predict passenger volume data and provide reference value for decision-making.
作者
陆奕
郭唐仪
LU Yi;GUO Tang-yi(School of Aautomation,Nanjing University of Science and Technology,Nanjing,Jiangsu 210094,China)
出处
《黑龙江交通科技》
2023年第1期128-131,共4页
Communications Science and Technology Heilongjiang