| 基于GM(1,1)-MLP神经网络组合模型的物流总额预测 |
| |
| 引用本文： | 张乐,汪传旭.基于GM(1,1)-MLP神经网络组合模型的物流总额预测[J].上海海事大学学报,2018,39(4):58-62. |
| |
| 作者姓名： | 张乐 汪传旭 |
| |
| 作者单位： | 上海海事大学经济管理学院,上海海事大学经济管理学院 |
| |
| 基金项目： | 国家自然科学基金（71373157） |
| |
| 摘 要： | 由于传统的基于GM(1,1)的物流总额预测方法需假设其他因素变化对物流总额无影响,给预测结果带来较大误差,本文采用GM(1,1)-MLP神经网络组合模型对我国未来物流总额进行预测。将组合模型与GM(1,1)的2012—2016年物流总额拟合结果进行比较,发现组合模型的预测平均误差仅为2. 3%,远低于GM(1,1)的预测平均误差(25. 2%),精准度大大提高,可以被有效应用于我国未来的物流总额预测。
|
| 关 键 词： | 物流总额预测 GM(1 1) MLP神经网络 |
| 收稿时间： | 2017/11/24 0:00:00 |
| 修稿时间： | 2018/2/7 0:00:00 |
Total logistics amount forecasting based on GM(1,1)-MLP neural network combination model |
| |
| Institution: | Economics & Management College, Shanghai Maritime Univeristy,Economics & Management College, Shanghai Maritime Univeristy |
| |
| Abstract: | Because the traditional GM(1,1) based forecasting method for the total logistics amount needs to assume that the change of other factors has no effect on the total logistics amount, which results in obvious errors to prediction results, the GM(1,1) MLP neural network combination model is adopted to forecast the future total logistics amount of China. The fitting results of total logistic amount in 2012 2016 obtained by the combination model and GM(1,1) are compared, which shows that the average prediction error of the combination model is only 2.3%, far lower than that obtained using GM(1,1) alone (25.2%), and the accuracy is greatly improved. It can be effectively applied to the forecast of the future logistics amount of China. |
| |
| Keywords: | total logistics amount forecasting GM(1 1) MLP neural network |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《上海海事大学学报》浏览原始摘要信息 |
| 点击此处可从《上海海事大学学报》下载免费的PDF全文 |
|
