基于改进IOWA算子的城市物流需求预测
首发时间:2020-08-26
摘要:为了进一步提高城市物流需求量相关预测模型的适应性和泛化能力,本文分别采用了BP神经网络、支持向量机、多元非线性回归构建单项预测模型,通过Shapley和MEM分别构建组合预测模型,并在此基础上结合改进的IOWA算子,得到混合预测模型。以泉州市次年货运量的预测为实例,对混合预测模型与其他预测模型进行对比,结果表明该混合预测模型预测精准度更高,相较于两个组合预测模型分别提升1.18%、0.47%。
关键词: 物流工程 混合预测模型 改进IOWA算子 次年货运量预测 最大熵 Shapley值
For information in English, please click here
Urban logistics demand forecast based on improved IOWA operator
Abstract:In order to further provide the generalization ability of the freight volume-related forecasting model, this article uses BP neural network, support vector machine, and multiple nonlinear regression to construct a single-item forecasting model, and builds a combined forecasting model through Shapley and MEM, and combines on this basis. The improved IOWA operator results in a hybrid prediction model. Taking the prediction of the next year's freight volume in Quanzhou as an example, the hybrid prediction model is compared with other prediction models. The results show that the hybrid prediction model has higher prediction accuracy, which is 1.18% and 0.47% higher than the two combined prediction models.
Keywords: Logistics Engineering Mixed Forecast Model Improved IOWA Operator Forecast of Freight Volume in Next Year Maximum Entropy Shapley Value
引用
No.****
动态公开评议
共计0人参与
勘误表
基于改进IOWA算子的城市物流需求预测
评论
全部评论