首页|基于Adam-LSTM的车用汽油价格预测

基于Adam-LSTM的车用汽油价格预测

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
汽油价格预测的国内外以往研究主要采用传统统计模型和简单神经网络模型,无法处理长期依赖性问题,而深度学习的长短期记忆(LSTM)神经网络模型能够有效处理成品油价格数据中的长期依赖问题,因此建立基于Adam(自适应矩估计)-LSTM人工神经网络模型,对四川地区 89 号、92 号、95 号汽油价格数据进行预测.首先对数据进行滑动平均处理,然后通过均方根误差(RMSE)、平均绝对百分比误差(MAPE)等指标评价模型的预测效果,同时通过DM(Diebold-Mariano)检验比较 LSTM模型与另外 4 种模型的预测效果,最终发现 LSTM对于该数据的预测效果最优.
Vehicle Gasoline Price Prediction Based on Adam-LSTM
Previous domestic and foreign research on gasoline price forecasting mainly use traditional statistical models and simple neural network models,which could not handle long-term dependency issues.However,the deep learning long short term memory(LSTM)neural network model can effectively handle the long-term dependency issues in crude oil price data.Therefore,an Adam(adaptive moment estimation)-LSTM artificial neural network model is established to predict the prices of 89,92,and 95 gasoline in the Sichuan region.The data is processed with moving averages,and then the model's predictive performance is evaluated using indicators such as root-mean-square error(RMSE)and average absolute percentage error(MAPE).Additionally,the predictive performance of the LSTM model is compared with four other models using DM(Diebold-Mariano)tests,ultimately it is found that the LSTM model has the best predictive performance for this data.

prediction of automotive gasoline pricesadaptive moment estimation(Adam)algorithmlong short term memory(LSTM)networkmoving average methodDM(Diebold-Mariano)test

于海洋、郭新旸

展开 >

西南石油大学理学院,成都 610500

车用汽油价格预测 自适应矩估计(Adam)算法 长短期记忆(LSTM)网络 滑动平均法 DM(Diebold-Mariano)检验

2023年四川省大学生创新创业训练计划

S202310615247

2024

科技和产业
中国技术经济学会

科技和产业

影响因子：0.361
ISSN：1671-1807
年,卷(期)：2024.24(15)