灰色多元回归模型在港口吞吐量预测中的应用

Grey Multiple Regression Model Applied for Port Throughput Forecasting

利用灰色模型和多元回归模型建立3种灰色多元回归模型:并联型、串联型和嵌入型。将各模型用于上海港吞吐量预测,可得结论:串联模型和嵌入模型是利用灰色理论对多元回归模型的改进,其能弱化原始数据的随机型,提高模型预测精度;并联型模型本质上是组合模型,能综合多种信息,预测具有非劣性,具有一定的实用价值。

Using grey model and multiple regression model, three kinds of grey multiple regression models including parallel model, series model and inlaid model were set up. Applying the models for Shanghai throughput forecasting, the conclusion was educed. Using the grey theory to improve multiple regression model, the series model and inlaid model can weaken the randomicity of original data and improve forecasting precision. As a combinatorial and noninferior model, the parallel model can synthesize much information and has a practical value.