两种组合预测模型在安徽货运量预测研究中的比较

摘要

文中首先采用单项预测方法，GM（1，1）模型和三次指数平滑模型对安徽货运量进行预算，然后在两者的基础上提出了IOWA算子预测模型和采用熵值法来确定组合预测模型，将以上两个单项预测模型在组合模型中赋予一定的权重，以实际量和权重的结合作为准则构造新的组合预测模型。以安徽省近年来货运量为数据，对2007－2012年货运量进行研究，结果表明基于IOWA算子的组合预测模型和基于熵权组合预测模型精度高于两种单项预测方法精度。最后，将两种组合预测模型进行对比，充分证明了IOWA算子模型在货运量预测上的有效性和科学性。%This paper uses the individual’s prediction method,GM (1 ,1 )model and three time exponential smoothing model to estimate Anhui cargo,then the two proposed based on the IOWA Operator predictive models and using entropy method to determine the combination forecasting model,the above two individual forecasting model gives a certain weight in the portfolio model,the actual volume and weight combination forecasting model with the new structure as a guideline.In recent years,Anhui Province cargo data for 2007-2012 cargo study results indicate that a combination forecasting model IOWA operators and two single precision forecasting method based on entropy weight combination forecasting model based on higher accuracy.Finally,the two combined forecasting model comparison,fully proved effective and scientific IOWA operator model in freight forecasting.